Any idea why an "A" author comes last in the bibliography

joehill

Hello,

I'm working on a paper now and for some reason "Abu-Zahra" is placed last in the bibliography. I'm using my own slightly modified version of the Chicago author-date style, although my modified version doesn't affect author names and as you can see here everything's alphabetical except this one case. I'd attach the file but I'm not sure how to do that. Any ideas?

Thanks,

Joe

Abdul-Rauf, Muhammad. 1979. The Islamic View of Women and the Family. 2nd ed. New York: Robert Speller.
Amin, Qasim. [1899] 1992. “The Liberation of Women.” In The Liberation of Women and The New Woman: Two Documents in the History of Egyptian Feminism. Cairo: American University in Cairo Press.
Antoun, Richard T. 1968. “On the Modesty of Women in Arab Muslim Villages: A Study in the Accommodation of Traditions.” American Anthropologist 70 (4): 671–697. doi:10.1525/aa.1968.70.4.02a00010.
[...]
Schulz, Dorothea E. 2003. “‘Charisma and Brotherhood’ Revisited: Mass-mediated Forms of Spirituality in Urban Mali.” Journal of Religion in Africa 33 (2): 146–171.
———. 2006. “Promises of (Im)mediate Salvation: Islam, Broadcast Media, and the Remaking of Religious Experience in Mali.” American Ethnologist 33 (2): 210–229.
[...]
Weber, Max. [1922] 1978. Economy and Society: An Outline of Interpretive Sociology. Berkeley: University of California Press.
Abu-Zahra, Nadia M. 1970. “‘On the Modesty of Women in Arab Muslim Villages’: A Reply.” American Anthropologist 72 (5): 1079–1088. doi:10.1525/aa.1970.72.5.02a00080.

adamsmith

It's the "Abu" part that Zotero ignores for alphabetization. There's a long thread about this that I haven't been following and don't have the link to atm - Gracile or fbennett will know where to point you.

joehill

I see. So it's considering Abu- to be equivalent to van or de?

joehill

If so, I think that's the wrong way to do it. If "Abu" is part of the last name field, it's always considered part of the name, not as a dispensable prefix. I don't think anyone familiar with Arabic names would think of looking for Abu-Zahra under Z or Abu-Lughod under L, or even non-hyphenated Abu Mazen under M. (And neither would anyone split it up into Mazen, ... Abu or Zahra, Nadia Abu-.)

adamsmith

I see. So it's considering Abu- to be equivalent to van or de?

not even - van and de are sorted under de and v in CMoS. This needs to go into the appropriate thread, I know too little about whether this is based on a concious decision or a bug to be of any help.

joehill

Well that's strange. I've been looking for appropriate threads but haven't found anything so far. I came across one that discusses non-Western naming conventions--
https://forums.zotero.org/discussion/2875/ -- but "Abu" in this conversation is mentioned as part of the last name that gets alphabetized. Another discusses the particle "al-" or "al " https://forums.zotero.org/discussion/28457/arabic-names-with-the-particle-al/ but doesn't mention "Abu". I've yet to come across any thread with a suggestion that Abu- should not go under A.

fbennett

@joehill

Thanks for this feedback. The change that's causing this behaviour went into the processor code on March 23; if it's too aggressive I'll dial the logic back to be more precise. Is "al-" the only hyphenated prefix that should be treated as a particle?

joehill

Yes that would be my preference. Abu- and even separated Abu are both considered part of the same name, as is Ibn if it's treated as part of a last name. So I would only treat the variants of Al (Al, al, Al-, al-), if any, in this way. (In Egypt they Romanize it as "El" but in that case they consider the Romanized name to be all one name, so ElBaradei goes under E, whereas if there were a medieval scholar publishing in Arabic under the same name it would be Al-Baradᶜī and would probably go under B.) That's just my opinion of where I would expect to look in the bibliography for a name--not sure if there's an official style guide anywhere about that.

adamsmith

what joehill says is exactly in line with CMoS:
http://www.chicagomanualofstyle.org/16/ch16/ch16_sec076.html

Arabic surnames prefixed by al or el (the) are alphabetized under the element following the particle; the article is treated like de in French names.

Hakim, Tawfiq al-
Jamal, Muhammad Hamid al-

Names beginning with Abu, Abd, and Ibn, elements as integral to the names as Mc or Fitz, are alphabetized under those elements.

Abu Zafar Nadvi, Syed
Ibn Saud, Abdul Aziz

(this is for indexing, but bibliographies should follow the same rules)

joehill

Right I was curious and just looked this up too. Thanks for checking.

joehill

But one question I have is, Chicago shows the al- coming at the end of the name (Hakim, Tawfiq al-) in the same way it shows "de" and other particles coming at the end (Beauvoir, Simone de; Keere, Pieter van den). Yet it specifies including the al- with the last name when citing it (al-Hakim). Is there a way to get Zotero to do that (attach the al- in citations, place it at the end in bibliographies)?

So far this issue hasn't come up with me for publishers (although I'm sure the Abu- issue would)--I'm just wondering whether it's possible since that's technically what Chicago asks for.

Gracile

[BTW, we really need a guide of good Zotero/CSL practices on this topic... I've worked on this quite a bit but I'm still confused and I always wonder in which field (first / last) I should store the dropping-particle..!]

Rintze

@Gracile: in writing the CSL specification, I assumed that the author names are supplied as parsed metadata. I would therefore prefer to keep the guidelines on parsing two-field names on the level of the reference manager.

Gracile

Rintze: I'm not sure I understand. I wrote "CSL" because this seems highly, i.e. entirely, connected to the way the CSL proc (citeproc-js here) handles that. But I was thinking about an how-to for Zotero. Did I miss something ?

Rintze

I just wanted to point out that there can be documentation on three levels: that of CSL, that of the CSL processor (citeproc-js), and that of the program that uses the CSL processor (Zotero).

When writing the CSL specification, I tried to limit it to describing how CSL styles should be interpreted. The spec doesn't touch on things that can't be controlled through CSL, such as parsing of "raw" dates into date parts, and of two-field names into name parts. While we should have guidelines for storing metadata in Zotero, that should be Zotero documentation, not CSL documentation.

Gracile

Ok. Thanks for the clarification. I think that's what I had (confusingly) in mind ;-)

joehill

Is there any workaround for reordering names while I'm waiting for Zotero to be updated to order names according to this discussion, or am I stuck with Zotero's automatic ordering?

DWL-SDCA

This is a problem on many levels. A few days ago I imported three articles by the same lead author from three consecutive issues of the same journal. In one, the name particle was after a comma in the first name field. In another it was included as a prefix in the Lastname field. In the 3rd it was in the last name field but it was cased differently. The case differences were in the metadata -and- on the printed page.

joehill

I'm just wondering if anyone has looked at rolling back this change so that Abu-Zahra comes near the beginning rather than the end of a bibliography? Is the only workaround now to edit the bibliography by hand just before we send an article out? Or is there a way to custom-order the bibliography?

adamsmith

This probably just slipped through fbennett's to do list. This should be a quick fix--Frank?

fbennett

Yep. The heuristics applied to particles by the processor are too loose to be satisfactory. I was fretting over better solutions, didn't hit on anything solid, and the issue slipped away.

Around the time of @joehill's first post, I received a nice list of name particles by category from Charles Parnot of Papers. If we use that, parsing can be made smarter and easier to control - and we might even be able to make parsing indifferent to field, addressing Gracile's concern over confusion above.

I'll work something up and release it via the patch plugin, so people with challenging libraries can test it before it's deployed in clients.

fbennett

I've opened a thread for feedback on the new name parser design here.

lancioni

I think is still unresolved on this issue. If I enter the "last name" of an author as al-Kindi, it still is sorted under 'A' (and not under 'K', as it should be). On the other hand, as I include al- as the last part of the "first name", it is space-separated from Kindi. I am using the full note Chicago style under MS-Word.

Gracile

In a couple of hours, Zotero 4.0.27.5 will be released. It includes some important changes concerning names and particles. You can read fbennett's comment here for the new general logic.

With zotero 4.0.27.5, "al-" is always demoted to the end: "Hakim, Tawfiq al-". Is that right for you ?

akateman

From the various threads on this topic, I get that name particles are quite a complicated matter, but I hope that there's a fix to my particular problem anyway.

It is related to this question of joehill asked above (whom I totally second in his complaint that Abu- in Arabic should not be dropped and should not in this sense be treated as a particle):
"But one question I have is, Chicago shows the al- coming at the end of the name (Hakim, Tawfiq al-) in the same way it shows "de" and other particles coming at the end (Beauvoir, Simone de; Keere, Pieter van den). Yet it specifies including the al- with the last name when citing it (al-Hakim). Is there a way to get Zotero to do that (attach the al- in citations, place it at the end in bibliographies)?"

My situation:

I put name particles (in particular "al-" and "van" or "van der") at the end of the "first name" box without a comma separating it from the actual first name.

My problems mainly occur with shortened citations:
- in shortened citations al- is dropped while Chicago suggests it should be kept in citations, as joehill described
- In contrast: "van der" is not dropped, yet not capitalized (as I think is preferable in Dutch names such as these, but should not be the case in a non-shortened citation: i.e. Van der Veer in a shortened citation and Peter van der Veer in a full citation)
- This seems inconsistent of Zotero for me, but probably there's a logic to it escaping me.

Then in the bibliography:
- the "van der"-name is alphabetized according to "van der" (according to its particle) which is really incorrect

I use Chicago Full Notes in Zotero (not stand-alone) 4.0.27.5 and Zotero Word for Windows Integration 3.1.20.

I hope someone has any idea to fix it. Thanks a million!

Gracile

Then in the bibliography:
- the "van der"-name is alphabetized according to "van der" (according to its particle) which is really incorrect

Edit your style, line 2, change the value of "demote-non-dropping-particle" to “sort-only” or “display-and-sort” instead of “never”.
https://www.zotero.org/support/dev/citation_styles/style_editing_step-by-step

nickbart

“al-” (and its variants, el-, ad-, as-, an-, at-, az-, etc.) are non-dropping particles, i.e., “al-Hakim” has to appear when used in the text itself, though the “al-” may be dropped for sorting and/or display in the bibliography (see Chicago Manual, 16, 8.14 “Arabic names” and 16.76 “Indexing Arabic names”).

As far as I can see, citeproc-js seems to be parsing "al-", incorrectly, as a dropping particle, irrespective of whether entered in the first or the last name field.

It would be great if citeproc-js’s parsing of “al-” and all of its variants could be fixed.

And a gentle reminder: Full documentation of citeproc-js’s name parsing algorithm would be much appreciated, not in the least for debugging cases such as this one.

fbennett

Re al-, I'll take a look, thanks for the heads-up.

Re documentation, so long as I'm working solo, reimplementation of the MLZ changesets for the new Zotero database methods will have to come first, because I have students and immediate colleagues that rely on it.

fbennett

“al-” (and its variants, el-, ad-, as-, an-, at-, az-, etc.)

Does anything specific come after "etc."?

nickbart

This is a reasonably complete list (a/e + sun letters, in the most common transcriptions):

at-, ath-, aṯ-, ad-, adh-, aḏ-, ar-, az-, as-, ash-, aš-, aṣ-, aḍ-, aṭ-, aẓ-, al-, an-,

At-, Ath-, Aṯ-, Ad-, Adh-, Aḏ-, Ar-, Az-, As-, Ash-, Aš-, Aṣ-, Aḍ-, Aṭ-, Aẓ-, Al-, An-,

et-, eth-, eṯ-, ed-, edh-, eḏ-, er-, ez-, es-, esh-, eš-, eṣ-, eḍ-, eṭ-, eẓ-, el-, en-,

Et-, Eth-, Eṯ-, Ed-, Edh-, Eḏ-, Er-, Ez-, Es-, Esh-, Eš-, Eṣ-, Eḍ-, Eṭ-, Eẓ-, El-, En-.

fbennett

Yeah, I probably wouldn't have picked those up.