5

When referencing previous results, should one cite the original paper(s) or a recent comprehensive monograph?

Specifically, in my area of interest (a relatively recent branch of mathematics), in the last ten years, several monographs have been published (by different authors) which cover (virtually) all "classical" results and most of contemporary developements organizing them, putting them in their context, and offering extensive references to the original papers.

To clarify my position: for the sake of "culture", I've read a few of the original papers in which the results I use most often appear for the first time; however, for the benefit of the reader, I'd rather cite only the most recent (and most comprehensive) source (and possibly point the reader to the additional references therein); in my department, there are researchers following either policies.

einpoklum
  • 39,047
  • 6
  • 75
  • 192
  • please someone relate this to this question: http://academia.stackexchange.com/questions/65892/citing-old-or-new-papers-of-same-work – beta Apr 21 '16 at 15:33
  • @beta: that question is somewhat related, but is radically different. –  Apr 21 '16 at 16:06

2 Answers2

10

Why not cite both?

The following theorem is due to Doe [37]; see also [28] for a modern exposition and further background.

Here [37] might be Doe's original paper, and [28] the recent monograph.

Math journals rarely have page limits, so there's no disadvantage to including more citations. This way you credit the original author and also point the reader to the resource you think will be most helpful.

Nate Eldredge
  • 133,015
  • 44
  • 379
  • 480
  • It may seem a little redundant, but I agree that it is still the best choice. Thank you. –  Apr 21 '16 at 16:05
  • 1
    This is pretty much also what I do with a lot of results, at least the ones I state in full. When I only cite a result as part of a proof, it would become too cumbersome to cite more than one source (I am not very consistent with which I choose to cite in these cases). – Tobias Kildetoft Apr 21 '16 at 17:25
1

The point of citations is to allow your work to be repeated. Therefore you must cite the materials you used. Consider if the recent monograph [28] had an error: if you cited the original [37] then that would be most unfair to people repeating you work. In the worst case that annoyed person attempting to repeat your work might publish a letter pointing out that your results could not have been acheived from your citations. That's a hair's breadth from being accused of misconduct.

However you also need to be fair to your reader. Imagine a reader who has tracked down hard-to-find monograph [28], painfully translated the relevant passages, and found it was merely an exposition of Doe [37].

You must cite the monograph. You may also cite the original paper.

Anon
  • 11
  • 1