To my eye, the contents of Crelle's Journal (Journal für die reine und angewandte Mathematik) around 1908 (not by chance the year Crelle's published Emmy Noether's dissertation) look very old-fashioned compared to those of the Mathematische Annalen in the same year. There are articles by currently well known mathematicians of course but not like in the Annalen.

Are there any authoritative statements or reasonably objective measures I could use to check my impression?

  • $\begingroup$ There are no objective measures for comparison of mathematical journals. $\endgroup$ Oct 12, 2016 at 1:43

1 Answer 1


MathSciNet has Journal Citation Search feature that might be useful. It lets you select a citing year (from 2000 to 2015 currently) and returns the histogram of total citations listed in MathSciNet from that year to past years (back to 1900), among other things.

With 2015 as the citing year it shows for Journal für die Reine und Angewandte Mathematik 13 citations to 1901-1904, 29 to 1905-1909, and 26 to 1910-1914. For Mathematische Annalen it shows 34, 45, and 111, respectively. For the citing year 2010 it is even more definitive: for Journal für die Reine und Angewandte Mathematik there is nothing to 1901-1908, 7 to 1909, and 1 to 1910-1914; for Mathematische Annalen there are 18 to 1901-1904, 52 to 1905-1909, and 71 to 1910-1914.

So Mathematische Annalen does appear increasingly more "modern" in those years, at least by these measures. You can try different citation years, and with extra work probably narrow it down to 1908 specifically, although I suspect that would require tracking individual citations.

P.S. Search links require MathSciNet access, and may be slow even with that.

  • 3
    $\begingroup$ One has to divide on the number of papers: otherwise these numbers may reflect just the number of papers published. Also citations in the 2000s of the paper published in the early 1900s do not really reflect their value. The typical life span of a math paper is about 40 years. $\endgroup$ Oct 12, 2016 at 21:10
  • $\begingroup$ Modern relevance and value are not necessarily the same (I suspect that "value" is a much more relative and loaded notion), and the fact that they are still cited after this long suggests relevance. Papers in highly selective journals are likely to outperform the average. But yes, it would be more reliable to average citations over a longer period than 2000-2015, if MathSciNet permitted :) $\endgroup$
    – Conifold
    Oct 13, 2016 at 21:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.