Journal of Molecular Endocrinology sorting

aurimas · August 23, 2012

I'm trying to follow the following rules for sorting the reference list:

List references in alphabetical order. Give articles by the same author in the order:

Single author
Two authors alphabetically according to the name of the second author
Three or more authors chronologically, with a, b and c etc. for articles published in the same year, in the order in which they are cited in the text.

I was trying to hack some things using et-al-min etc, but found out that:

A) et-al suffix and prefix does not work even though it's in the specification
B) even though et-al is ignored during sorting, adding text after the names field or adding a suffix to the names field, appears to result in an extra space before the prefix/space because something like "Beato et al. ZZZZ" sorts before "Beato ZZZZ"

Any ideas how to code the sorting for the rules above?

adamsmith · August 23, 2012

that should work with names-min and names-use-first, no?
http://citationstyles.org/downloads/specification.html#sorting

aurimas · August 23, 2012

I don't think so, but I could be wrong. The four references discussed below are in this RDF file.

Consider entries:
1)"Beato M, Adams P & Schütz G 1995 Steroid hormone receptors..."
2)"Beato M, Brändle W, Biesewig D & Sekeris CE 1970 On the mechanism of hormone action..."
3)"Beato M & Schütz G 1995 Fictional a..."
4)"Beato M 1989 Gene regulation by steroid hormones..."

Based on the description above, I think the ordering should be:

4)
3)
2)
1)

I tried with names-min and names-use-first, which I assume you meant 3 and 2 (or 1) respectively, but I don't see how (3) can be pushed above (1) and (2)

Also, I realized just now that (4) was being sorted last because that entry has the full first name instead of just initials. I can see how that can easily be fixed in a sorting macro though.

Anyway, still would like some insight into making 2 author names sort above 3 author names, but 3 author names sort only based on first author and year.

I understand the logic behind this sorting as trying to sort names essentially like they appear in text, except that "et al." should sort to the end.

I hope that makes sense

Edit: Fixed "I realized just now that (1) was being sorted last" to (4)

Rintze · August 29, 2012

It's somewhat complicated, but you probably want to sort by the number of authors, which can be done using the "count" form on the cs:name element.

See e.g. the sorting logic in https://github.com/citation-style-language/styles/blob/master/american-geophysical-union.csl and the description of the "form" attribute at http://citationstyles.org/downloads/specification.html#name

Rintze · September 3, 2012

@aurimas, do you need affixes on cs:et-al?

aurimas · September 3, 2012

See e.g. the sorting logic in https://github.com/citation-style-language/styles/blob/master/american-geophysical-union.csl

Thanks!! That example was super helpful. The sorting works exactly as required now.

@aurimas, do you need affixes on cs:et-al?

I do not. I was trying to hack something up to get things sorted correctly, but it seems that there are better ways to do it. I don't think it would have helped either way, since "et al." is ignored for sorting.

Nevertheless, did I misunderstand the documentation, or are you supposed to be able to assign affixes for et al? If so, I was not able to do it.

adamsmith · September 3, 2012

Nevertheless, did I misunderstand the documentation, or are you supposed to be able to assign affixes for et al? If so, I was not able to do it.

this was supposed to have been possible but has been removed with the CSL 1.0.1 update that just went live
http://citationstyles.org/downloads/release-notes-csl101.html#affixes-on-cs-et-al

which means we would now also accept the delimiter-precedes-et-al attribute that I believe you wanted?

aurimas · September 3, 2012

Great! I updated the style.

Rintze · September 4, 2012

For a bit more background on affixes on cs:et-al, see https://github.com/citation-style-language/schema/issues/101. I think you're the first person to complain in over two years that they don't work, which is a pretty good argument that we don't need them.

Gracile · September 4, 2012

@Rintze:
A style like this one requires brackets: [et al.]. Since it seems to be a general requirement, the term has been redefined in cs:locale. Affixes are not necessary.
I can't imagine a style where the brackets would only be required in certain cases...

Rintze · September 4, 2012

If the need arises, we can always reallow them in the future. Not allowing them has the advantage that style authors cannot mistakenly assume that affixes control the delimiter before "et al.", though (there are currently a few styles that try to do exactly that, e.g. define a space as the prefix, where they probably want to use delimiter-precedes-et-al="never").