Neutral Citation Sorting
I’m compiling a list of Supreme Court decisions and one of the data I’m collecting for each is the neutral citation. I’ve just noticed something that should have been obvious to me from the first: when using alpha-numeric sorting, neutral citations don’t sort properly by the ordinal number (i.e. the number that is located where the page number would be in a traditional citation.). This is because the ordinal numbers don’t all have three places. Thus 2006 SCC 3 (the citation for Young v. Bella) winds up along with the pages numbered 30 etc. because the computer can’t supply the missing 0’s in the ordinal number.
I realize that this is a very minor issue, but it does reveal I think (unless I’m missing something) that the committee that created neutral citations hasn’t thought about computerized sorting, which is quite useful and likely to become more important as the legal profession moves even further into the 20th century.
Simon
You can build special sorts for such compound references (in this case year – series number) into a system. For example, Windows XP will correctly sort a file named 2006 SCC 4 betwen 2006 SCC 3 and 2006 SCC 30.
Specifying that a citation number should b padded with zeroes to x digits so that the citation can be sorted as a string would cause more problems than it solves.
Nick
Not sure I agree with you Nick. Your way you make all users who want to sort modify their systems, instead of modifying the main single source. Besides, some systems and programs can’t be altered either easily or at all.
It would depend, I suppose, on what sort of problems moving to a 001 system would produce. Can you give me an example?
I don’t wish to spark an extended debate on data processing, but for me the nub of it is that 001 is not a natural number (in fact it’s the appellation of a predecessor of James Bond!) and this limits how it can be stored and used as a discrete item of data. Also, should it be 001 or 0001 or even 00001? citers will get it wrong.
Clearly lots of legacy sytems currently store a cite as a single data item, rather than as 3 discrete items which is much more useful and flexible. Padding the number solves only the string-sort-by-number issue; for anything more the string needs to be parsed and the 3 items processed.
So, if we’re looking to the future, I think legacy systems will have to change, either to enable parsing of such data or to restructure the data. Remember Y2K?
The Canadian Citation Committee (CCC) did think about computerized sorting, but for the reasons stated by Nick Holmes as well as for better legibility, this issue was adressed not at the level of the citation itself, but rather at the level of file naming, where it really belongs. See par 25 of the CCC’s Guide to the Uniform Preparation of Judgments, which states that courts may wish to zero-fill the filename of the judgment to facilitate sorting: http://www.lexum.umontreal.ca/ccc-ccr/guide/guide.prep_en.html#_Toc20214410
Thanks, Frédéric. And my apologies for having done my research in a hurry. The reference to “zero-filling” makes me realize I wasn’t entirely nuts.