Submission and publication of mathematics papers

A proposal was aired in April 2009 that academics in Science and Engineering at my university should be asked (or forced) to submit papers to one of a list of “approved” journals. This proposal has caused some disquiet: it goes far beyond any “management” of research previously attempted or suggested. The purpose of this paper is to point out the dangers and unwisdom in such an approach, and to examine whether its objectives could be better met by cooperation rather than confrontation.

Since then, things have changed a bit; the Funding Council have backtracked on their commitment to bibliometrics, and ensured us that the next Research Assessment Exercise (which will go by the trendier title of “Research Excellence Framework”) will be done in the same way as the last one, by experts on panels actually reading papers, though they will have bibliometrics available to them to use if they wish.

I am publishing my response to this proposal in the hope that it may be useful
to colleagues elsewhere who may be facing the same difficulties.

Choosing a journal

There are many factors to consider in choosing a journal for a paper, beyond the aim of getting it in the “best” journal. It must be clearly noted that we do not do research, or write papers, for the purpose of being assessed. Coupled with this is the fact that we know better than any research manager what the appropriate place for our papers is.

• The paper may be a survey. Such papers can be among the most-cited of all, which should increase their importance to the bean-counters as we move from evaluation of papers by experts to bibliometrics. Survey papers may be in top journals, but could be in journals with a preference for such papers, or in special one-off volumes (such as the Surveys in Combinatorics volumes produced for the British Combinatorial Conferences).
• The paper may be submitted to a volume celebrating or commemorating a mathematician. If that mathematician is or was a close colleague or friend of mine, I want to put a good paper in the volume, irrespective of the journal publishing it.
• The paper may be something else entirely, such as a review, in which case most research journals are not at all the appropriate place!
• I may want to target my paper to a particular audience (see below).
• Journals evolve and may become more or less prestigious. Also, the interests of mathematicians evolve, and a list set in stone will not represent our interests for long.
• Costs of journals vary enormously. Many libraries cannot afford to take the more expensive journals. I may not want to submit to a journal not in my library, no matter what its impact factor.

Mathematics journals fall into two broad classes: general journals (such as those of the London, European or American Mathematical Societies, or others such as Inventiones, Annals or Advances); and specialised journals, which themselves vary in their degree of specialisation. For example, in the field of Combinatorics, there are journals such as J. Combinatorial Theory, Discrete Math., European J. Combinatorics or Electronic J. Combinatorics, which cover the whole field; there are more specialised journals such as J. Algebraic Combinatorics, J. Graph Theory, J. Combinatorial Designs; and there are journals which interface with other fields such as Designs, Codes and Cryptography or Combinatorics, Probability and Computing. If I want my paper to be read by experts in a particular subject group, or to attract attention of researchers in a related field, this will influence my choice of journal.

The Appendix to this document is a list of some journals to which two recent Fields medallists, Terence Tao and Wendelin Werner, have submitted papers. It would be difficult to argue that a journal attracting authors of this calibre is not good enough for ordinary mathematicians!

Bibliometrics

There are various reasons why bibliometric measures for science and engineering do not do very well for mathematics journals. For example, a recently published list of top mathematics journals had a very poor representation of those that we would consider the best, and was made up largely of journals in statistics and operational research. These journals would not appreciate submissions of papers in group theory!

Two factors contribute to this:

• Publication times in mathematics are very long. Three years from submission to publication is fairly typical. Furthermore, highly-regarded mathematics papers are still being cited twenty, or even a hundred, years after their publication. The citation index uses a window of one to two years from publication. The figures for mathematics papers are so low because it is very difficult to get a paper into print within two years of the appearance of a paper it cites.
• Impact factors vary greatly from one subject to another, and even from one area to another within a subject. Perhaps statistics journals score highly because they contain the occasional paper which is relevant to medicine. Pure mathematics has no such crutch. Any comparison of areas based on this measure will be flawed.

Appendix: Journals chosen by two recent Fields medallists

Terence Tao:

• Acta Math. (4)
• Adv. Appl. Clifford Alg. (1)
• Adv. Math. (2)
• Amer. J. Math. (8)
• Analysis and PDE (2)
• Ann. Inst. Fourier (1)
• Ann. IHP (1)
• Ann. Math. (6)
• Ann. Math. Studies (1)
• Ann. Probab. (1)
• Ann. Statistics (1)
• Appl. Numer. Harmon. Anal. (1)
• Bull. Amer. Math. Soc. (3)
• Combinatorica (1)
• Comb. Probab. Comput. (1)
• Comm. PDE (4)
• Comm. Math. Phys. (4)
• Comm. Pure Appl. Math. (3)
• Disc. Cont. Dynam. Systems (3)
• Duke Math. J. (7)
• Dynamics of PDE (4)
• Electronic J. Combinatorics (2)
• Electronic J. Diff. Eq. (3)
• Ergodic Theory Dynam Syst. (1)
• Forum Math. (1)
• Geom. and Funct. Anal. (10)
• IEEE Inform. Thy. Newsletter (1)
• IEEE Trans. Inform. Thy. (4)
• Indiana U. Math. J. (2)
• Int. Math. Res. Notes (2)
• J. Amer. Math. Soc. (10)
• J. Anal. Math. (1)
• J. d’Analyse de Jerusalem (3)
• J. Austral. Math. Soc. (3)
• J. Diff. Eq. (2)
• J. Europ. Math. Soc. (1)
• J. Funct. Anal. (6)
• J. Hyperbolic Diff. Eq. (2)
• J. Inst. Math. Jussieu (1)
• J. Math. Phys. (1)
• J. Royal Statist. Soc. (1)
• J. Th. Nombres Bordeaux (1)
• Math. Res. Letters (9)
• Mathematika (1)
• Math. Ann. (5)
• Math. Comp. (1)
• Math. Z. (2)
• New York J. Math. (3)
• Nonlinearity (1)
• Notices Amer. Math. Soc. (2)
• Pacific J. Math. (1)
• Proc. Amer. Math. Soc (2)
• Proc. Edinburgh Math. Soc. (1)
• Proc. London Math. Soc. (2)
• Quart. J. Math. (1)
• Quart. J. Pure Appl. Math (1)
• Random Struct. Alg. (3)
• Rev. Mat. Iberoamericana (4)
• SIAM J. Math. (2)
• Trans. Amer. Math. Soc.

Wendelin Werner

• Acta Math. (3)
• Ann. Fac. Science Toulouse (1)
• Ann. IHP (3)
• Ann. Math. (1)
• Ann. Probability (4)
• Bernoulli (1)
• Bull. London Math. Soc. (1)
• Comm. Math. Phys. (1)
• C. R. Acad. Sci. Paris (1)
• Electronic Comm. Probab. (4)
• J. Amer. Math. Soc. (2)
• J. Europ. Math. Soc. (1)
• J. Statist. Phys. (1)
• Math. Proc. Cambridge Philos. Soc. (1)
• Math. Res. Letters (2)
• Pramana J. Phys. (1)
• Probab. Surveys (1)
• Probab. Th. Related Fields (7)
• Progress Probab. (2)
• Stochastics (3)
• Stoch. Proc. Appl. (1)

About these ads