New Hart's Rules: The Oxford Style Guide 2nd edition (numbered)

Lagerfors · 2026-02-04T16:21:54+00:00

Hi,

I was trying to follow "Research Connection" (Oxford University Press) style guide by using the "New Hart's Rules: The Oxford Style Guide 2nd edition (numbered)" style in Zotero.

However, I found that the style was not numbering my citations in sequence in the text - but apparently randomly. I couldn't figure out the pattern. It must be a bug.

I found out that one sequence in the style code was doing this. In Zotero's Style Editor I modified the following section (very much at the bottom of the script):

I erased everything between and including and and saved the style under an amended name. From there on it worked correctly.

Maybe this is useful to someone else working with Oxford University Press journals.

adamsmith · 2026-02-04T16:34:52+00:00

I do suspect this is an error in the Hart's Rules numeric (CC @dunning ), but the journal doesn't follow that style anyway. They say they use Oxford SciMed, which we have separately available: https://www.zotero.org/styles?q=id:oxford-journals-scimed-numeric

Lagerfors · 2026-02-04T17:07:02+00:00

Hi Adam,
thanks a lot. I actually used https://www.zotero.org/styles?q=id:oxford-journals-scimed-numeric but the bibliography was rejected by the journal, mainly due to missing DOI links.

dunning · 2026-02-04T17:18:18+00:00

Yes, the two SciMed styles say they are from New Hart’s Rules, but they are so divergent from what is generally shown in the guide, and the OUP style sheets provide so little detail, that I never worked out what they really want. I’ll take another look to see if I can sort out the logic.

adamsmith · 2026-02-04T17:30:35+00:00

But the sorting in the New Hart's numeric is a bug, no?

As for SciMed... Frustrating.
They don't provide any guidance on the actual style and say they don't care ("Authors may format references in any readable style at submission") on the website... Oxford doesn't have a central SciMed guide online and the journals that do use it and have checklists all have journal articles without DOIs.
And to add insult to injury, Research Connection doesn't even include DOIs in published articles, so there's literally no way to tell.

dunning · 2026-02-04T18:21:53+00:00

Yes, the sorting bug is entirely my fault, which should be fixed here:

https://github.com/citation-style-language/styles/pull/8048

I suppose we should just add DOIs to SciMed if that's what they want in practice?