Friday 1 June 2012

Citation tracking

I know of no good way to quantify the value of one person's contribution to science, or indeed any other field of intellectual endeavour. Nevertheless, people often try to do so, and the vast majority of these efforts focus on citation counts. There are many suggestions as to what is the most reliable indicator: total citation count, average citations per paper, or something slightly more sophisticated, like the h-index. Needless to say, all of these have major flaws, and tend to favour scientists who have simply been around for a long time, or who work on popular topics, but I get the impression that such quantities are still sometimes used as an aid in decisions on hiring, tenure etc.

Such issues have been discussed exhaustively on the blogosphere and elsewhere (although I'm too lazy to gather any links), and I don't really have anything new to add. But I did want to point out that, as far as I can tell, accurate citation data often simply aren't available. In high energy physics, we are lucky to have INSPIRE, which does a remarkably good job of collecting and organising citations from the literature on particle physics, string theory, gravity, and related fields. This is helped enormously by the fact that the vast majority of such papers are posted on the arXiv. But even this isn't foolproof; some of my papers, for example, overlap with the interests of pure mathematicians, and have been cited by papers which aren't included in the INSPIRE database. Most of these (that I know of) have been picked up by Google scholar, but that service misses some citations which INSPIRE finds.

Here are my own papers, with citation information, on three different free services: INSPIRE, Google scholar, and ADS. You can see that things generally agree reasonably well, but in some cases there are large discrepancies. Since I don't know the source of these, it could be that other authors are affected much more dramatically. I have no idea what the situation is like in completely different fields.

It is nice to be able to keep track of one's citations (I think most people keep an eye on theirs), but there doesn't seem to be a reliable way to catch them all.

15 comments:

  1. Quantity of citations is not the same thing as quality of citations. Woit and Smolin have a somewhat negative view of string theory while Davies and Candelas have a rather more positive view of string theory. So far as I am aware, the terms "Rañada-Milgrom effect" and "Space Roar Profile Prediction" have never occurred in a refereed physics journal and perhaps never shall occur. I would welcome any comments from string theorists on the following:
    MILGROM DENIAL HYPOTHESIS: The main problem with M-theory is that M-theorists fail to realize that Milgrom is the Kepler of contemporary cosmology.

    ReplyDelete
  2. A citation is only as good as the person who makes the citation. What is the greatest book ever written? I think the answer might be Francis Crick’s “What Mad Pursuit”. According to Will Rogers, the two best ways for us to learn are by reading good books and by talking to people who are smarter than we are. James D. Watson admonishes us, “Never be the smartest person in the room.”
    Everyone is a role model in the sense that intelligent behavior shows us what to do and stupid behavior shows us what not to do. I have learned a lot from string theorists. The way that string theorists have ignored the criticisms of Philip Anderson, Sheldon Glashow, Robert Laughlin, and Burton Richter and the work of Milgrom, McGaugh, and Kroupa have lead me to the following ideas:
    Trifecta of Theoretical Stupidity:
    (1) Do not work with experimentalists.
    (2) Do not gain a broad and deep understanding of relevant, empirical evidence.
    (3) Ignore well-informed critics.

    ReplyDelete
  3. The two comments above can stand, David, but I deleted your other two long, nonsensical rants. The comments section of my modest little blog is not the place for you to lay out your thoughts at length.

    ReplyDelete
  4. Thanks for not deleting my first two posts.

    ReplyDelete
  5. What you call my nonsensical rants have 2 main points: Gerald Rosen has a PhD from Princeton with Valentine Bargmann as thesis advisor. The ideas of Koide, Rosen, and Brannen on preon theory might be embeddable into M-theory merely by assuming that superstrings have 3 distinct energy levels.

    ReplyDelete
    Replies
    1. Many people have PhDs in physics from many good institutions, with many good advisors. Far fewer people are excellent physicists.

      Superstrings have infinitely many energy levels, and it couldn't be any other way. Koide's formula is intriguing, but could well be a coincidence, because none of the attempts to embed it in a theoretical framework look at all convincing.

      In any case, all of this is off-topic.

      Delete
    2. Sorry by coming late to this off-topic :-)

      Let me mention that Koide formula for quarks, besides advancing a lot in the last year (see my prezi http://prezi.com/e2hba7tkygvj/koide-waterfall/ ), predates the one for leptons and was actually considered in models with discrete symmetry. The first triplet that fulfills Koide formula appears, I think, in "Quark masses and cabibbo angles", by Harari, Haut and Weyers. It is very particular because it sets one of the masses to zero, for the up quark, then just predicting a quotient between strange and down. It was later critiquised on the ground of asking for a very complicated Higgs sector, if it was to be built via yukawas and Higgses. But at least Scopus gives it 152 citations, not a bad metric. Better than Koide "A New Formula For The Cabibbo Angle And Composite Quarks And Leptons", which only gets 44 citations according INSPIRE.

      Delete
  6. "... quantity the value of one person's contribution ..." At least physics has experiments. What I mean by "superstrings have 3 distinct energy levels" involves an 11-dimensional superfluid with 3 energy=density levels coexisting or, perhaps, Bagger-Lambert-Gustavsson action with Koide-Brannen-Rosen symmetry.

    ReplyDelete
    Replies
    1. David, the logical way to get Koide relations from string theory (which is something that I'm investigating) would be through geometric symmetries which show up in the "yukawas" or "yukawa couplings" describing how strongly the Higgs field(s) couple to the fermions.

      anonymous because I don't want David Brown adding me to his list of enthusiasms and spamming a dozen blogs with my name...

      Delete
  7. Anonymous: I disagree. The way to do it, in my opinion, is to replace hbar, alphaprime by hbar, alphaprime1, alphaprime2, alphaprime3. I mean to say that the superstrings need to be under 3 different string tensions. If the yukawa couplings method works then it should be mathematically equivalent to Koide-Brannen-Rosen symmetry.

    ReplyDelete
  8. Anonymous: I hope you can successfully get Koide relations from string theory. If you can do this then you might not win the Nobel prize but you should at least get the Wolf Prize or the Japan Prize. My advice to you is to keep secret your great ideas until you get published in a refereed journal. The only reason that I am not feverishly attempting to embed preon theory into M-theory is that I am an incompetent mathematician, or, at best, a third-rate mathematician. If hundreds of string theorists work on embedding preon theory into M-theory, then I see that as good for me but perhaps bad for you.

    ReplyDelete
    Replies
    1. For string theory, preons would be a highly exotic approach. If you see Koide's own recent papers, nowadays he is developing a type of "flavon" model, in which the standard model fermions are elementary, but the yukawa couplings come from the vacuum expectation values of new Higgs-like fields. That should be capable of dovetailing with very mainstream string model-building, in which (as I understand it) flavons might be derived from combinations of CY moduli.

      Since Rhys's work is playing a central role in my own approach to the Koide problem, maybe I'm giving the game away by blabbing on his blog. If the answer does lie in the part of the string landscape that he and his collaborators investigate, who knows how quickly they might find it if they started to take the problem seriously? Then again, it's not as if heterotic yukawas are easy to calculate. So it could drag out for a long time.

      In any case, I'll be posting elsewhere soon about my own line of thought, and we'll see where it all leads.

      Delete
  9. In any case, good luck to Anonymous and Rhys Davies.

    ReplyDelete
  10. According to Wolfram Alpha,
    .728^(1/64) = .99503069...
    According to Wikipedia, 72.8 % is approximately the % of dark energy in the universe.
    http://en.wikipedia.org/wiki/Lambda-CDM_model
    According to Wikipedia, .995 is approximately the CKM CP-violating phase.
    http://en.wikipedia.org/wiki/Standard_Model
    Let us assume that Newton/Einstein gravitational theory is 100% correct (contrary to what Milgrom/McGaugh/Kroupa’s findings would seem to indicate). In that case there should a Milgrom field that explains Milgrom’s acceleration law and there should be a Nernst field that explains the resolution of the vacuum catastrophe. By relating the Nernst field, the Milgrom field, and the inflaton field to the Higgs field, string theorists might be able to explain the cosmological constant in terms of the Standard Model of particle physics. By elaborating on Koide’s work, i.e., embedding preon theory into M-theory, string theorists might be able to explain all the free parameters of the Standard Model.
    http://en.wikipedia.org/wiki/Yoshio_Koide

    ReplyDelete
  11. Is there an SO(64) analogue of Gauss's Theorema Egregium?

    ReplyDelete

You can use LaTeX code in your comments simply by enclosing it in dollar ($) signs.