citation-style-language / documentation

Citation Style Language documentation
http://citationstyles.org/
Creative Commons Attribution Share Alike 4.0 International
42 stars 21 forks source link

What are the rules for shortening a year? #64

Open customcommander opened 5 years ago

customcommander commented 5 years ago

Hi,

This is related to the behaviour of the form="short" attribute on a year's <date-part> element, e.g. <date-part name="year" form="short"/>. Cf date-part documentation

The above date-part element would render 2005 as 05.

Questions:

  1. Is keeping the last two numbers when shortening a year the only rule when shortening a year?
  2. Should years such as 1666, 452 or -734 be allowed to be shortened?
  3. Should only years in the 21st century be allowed to be shortened?
  4. Should shortening a year be allowed only if it doesn't create confusion? e.g. 00 can be the short version of 2000 or 1900.
bwiernik commented 3 years ago

Shortened years are primarily used in physical sciences styles, where references to older works isn't common, so I think that permitting shortened years on any positive (CE) 4-digit year (i.e., whenever CE/AD or BCE/BE wouldn't apply) is fine. If we drew a later line, such as 1950 or something, I'm not sure where that would appropriately be.

@denismaier @bdarcus ?

denismaier commented 3 years ago

Not really sure, but maybe "shorten years if current year - issued year < 100". I.e, in 2020 you'd shorten all dates beginning with 1921.

denismaier commented 3 years ago

Otoh, I think that's done mostly with label citation styles where I think this ambiguity is not so problematic as @bwiernik has pointed out.

bwiernik commented 3 years ago

I'm not sure how much we want to embed "current year" into CSL processing? In any event, I think it is a rare enough occurrence that we can just apply the same rules as BC/AD and call it a day.

denismaier commented 3 years ago

I'm not sure how much we want to embed "current year" into CSL processing? In any event, I think it is a rare enough occurrence that we can just apply the same rules as BC/AD and call it a day.

Probably correct. I'm just wondering whether this is actually used in styles? As said above, I think this should mostly be relevant for label citation styles, and I doubt there will be much for for this in author date or note styles. If so, while "Ein07" my be fairly common, we'll probably never see "Einstein 07" in an author date style.

Btw, I've checked what biblatex has to offer here: Searching the manual for "two digits" only leads to two results—both in the same paragraph—that discuss how labels in label citation styles are being constructed.

customcommander commented 3 years ago

Just to give you some context: as a pet project I'm writing a CSL processor so these are just questions about what it should do in situation X, Y or Z (as I wasn't too sure). It's not meant to be a bug report in the specs. Although if it highlights areas that could do with a bit of clarifications then bonus.

Thanks both for looking into this though :)

denismaier commented 3 years ago

Just to give you some context: as a pet project I'm writing a CSL processor so these are just questions about what it should do in situation X, Y or Z (as I wasn't too sure). It's not meant to be a bug report in the specs. Although if it highlights areas that could do with a bit of clarifications then bonus.

Thanks both for looking into this though :)

Cool. What is that for?

bwiernik commented 3 years ago

Okay, cool. Let's go with short years following era marker rules.