Slashdot Mirror
Top 100 Papers in Physics Ranked
Rob Carr writes "What do physicists care about most? Who are the greatest minds of our time? What physics papers have had the greatest impact? Sidney Redner attempts to answer that question by looking at the citations of all journals in the Physical Review Journals since 1893. He ranked the top 100 papers based on their 'impact': the number of citations times the average age of the citations. Einstein's Relativity papers, which were not in Physical Review journals, are the most stunning absence. 'Fan Favorites' are there - Einstein does make the list for the Einstein Podolsky Rosen paper. Feynman, Dirac, Bethe, Wheeler are on the list. Stephen Hawking does not make the list. Yet Nobel Prize winner Walter Kohn, who is virtually unknown to the general public, is an author on five of the 100 papers, including the top two and one of the top 15 'hot' papers. The paper goes into the statistics of the citations, a fascinating area in it's own right. Some papers make an immediate splash, while others might wait 50 years before their importance becomes apparent. The vast majority die a quick and quiet death. It's tempting to wonder if Redner's paper conclusively proves Sturgeon's Law."
Yeah, but it's just Phys Rev. A lot of cool stuff happens that never gets published in Phys Rev. Sometimes, it's a talk at a symposium that is published and makes a big splash.
Neither. Lot's of people have been using Walter Kohn's theory. The reason why he is at the top of the list is because of the sucess of density functional theory (DFT) first in condensed matter physics and then in chemistry. A goodly portion of the unclassified CPU power used my scientists around the world is probably dedicated to examining systems with DFT.
Essentially, there are two neat things about DFT. The first is that it proves that it is possible to fully describe the state of a bunch of electrons with the 4 dimensional spin density, rather than the normal 4N coordinates (where N is the number of electrons, 3 cartestians an a spin per electron). This, combined with Kohn-Sham theory results in a method of calculating electronic structure that formally scales and N^4, but gives answers often as accurate as N^5 and higher methods. Hence, Nobel Prize :)
for example, feynman no doubt did some great physics, but he gets much, MUCH greater recognition over two other guys who did the same work (tomonaga and schwinger, they shared the nobel prize)
You're correct that Feynman was a more dynamic speaker/teacher, etc. But I think it's a bit of a jump to say that that's the only reason why he gets more attention than Schwinger and Tomonaga. For starters, they didn't all do the same work, even on QED. It's true that all three arrived at equivalent formalisms for calculating amplitudes, but that's not the same as saying they did the same work. Have you thrown away Feynman diagrams and straightforward perturbation expansions and instead tried to do things the way Schwinger did? It's a bitch! As a famous quote of the time went, "Feynman shows you how to do it; Schwinger shows you that only he can do it." And that had a lot to do with the eventual predominance of Feynman's perspective, and thus his getting more recognition than Schwinger or Tomonaga.
Furthermore, while I can't speak to Tomonaga in this regard, Feynman made a major splash in a much broader spectrum of physical investigations than Schwinger did. The work on QED was simply one of many arguably Nobel-worthy accomplishments of his. That, too, contributes to his being paid more attention to than Schwinger and Tomonaga.
Of course, you could argue that these are only things that matter to the cognoscenti; they don't explain why Feynman is more recognized by the general public. But I would claim that contrary to what physicists, and geeks who like physics, think, the general public is pretty oblivious to physicists entirely. They've heard of Einstein; they might have heard of Hawking. That's pretty much it, though. We think of Feynman as famous; the average person on the street has never heard of him.
So while I would agree that Feynman's dynamic personality, excellence in presentation, etc., is important in the way he is remembered by those who are aware of him at all, at least equally important is the fact that he did a ton of amazing new physics.
"...the number of citations times the [average citation age]..."
It seems to me that this nullifies the comparison in some regards. If you rank by this number DEscending, you get a few old papers with a lot of citations... possibly just because they're old. If you rank by this number Ascending, you get just the newest papers without significant numbers of citations. It might be better to rank by either total numbers of citations or "the number of citations *divided* by the average citation age", and use a DEscending rank. This way, recent works get a 'fair' (or 'fairer') comparison against older works.
Let S_n = {nst+us+vt : s,t in Z \ {0}, u,v in {-1,1}}. For all n in Z where |n| > 2, Z \ S_n is infinite... right?
Well, sure, it is not going to win this guy a Nobel prize, but it is interesting. Maybe not "research" by many definitions of the word, but definitely interesting.
For example, while I am quite familiar with DFT and have read most (if not all) of the Kohn papers mentioned in the article, I would not have guessed he would have placed so high. But that is the neat thing. This paper shows how much physics and chemistry interact. Many of the other paper in this top 100 list are probably more cited in the chemistry literature than in physics (e.g. Carr-Parinello)
Oh, and am I the only one that chortled at the fact that this paper, which lists the 100 most cited papers, had only 26 references?
heh, nope.
and how about these?
Read before you cite!
cond-mat/0212043
We report a method of estimating what percentage of people who cited a paper had actually read it. The method is based on a stochastic modeling of the citation process that explains empirical studies of misprint distributions in citations (which we show follows a Zipf law). Our estimate is only about 20% of citers read the original.
Copied citations create renowned papers?
cond-mat/0305150
Recently we discovered (cond-mat/0212043) that the majority of scientific citations are copied from the lists of references used in other papers. Here we show that a model, in which a scientist picks three random papers, cites them,and also copies a quarter of their references accounts quantitatively for empirically observed citation distribution. Simple mathematical probability, not genius, can explain why some papers are cited a lot more than the other.
sPh
You can call the systematic study of just about anything, digging through different sources, "research." It's just that this is research on physics papers and their authors, not on physics itself.
The correlation between ignorance of statistics and using "correlation is not causation" as an argument is close to 1.
From a theory of Hohenberg and Kohn, approximation methods for treating an inhomogeneous system of interacting electrons are developed. These methods are exact for systems of slowly varying or high density. For the ground state, they lead to self-consistent equations analogous to the Hartree and Hartree-Fock equations, respectively. In these equations the exchange and correlation portions of the chemical potential of a uniform electron gas appear as additional effective potentials. (The exchange portion of our effective potential differs from that due to Slater by a factor of 23.) Electronic systems at finite temperatures and in magnetic fields are also treated by similar methods. An appendix deals with a further correction for systems with short-wavelength density oscillations.
I kinda sorta knew what they were talking about up until Hartree and Hartree-Fock. After that I have no idea. For most of these papers, you really do need some graduate level education to know what's going on..
I'm sure it would be neat to go through and read a lot of these papers, but it's going to be very hard and very slow and you are going to have to have tons of background material at your side just to get the most basic meaning from them. I can't imagine reading papers in my field and getting much meaning from them before I went to college.
IMHO, if your goal is learning, you'd be much better off with some good textbooks. I know a textbook isn't as glamorous as reading the most cited papers in physics, but you'll make way more progress towards learning your area. There are some really good textbooks out there in most fields. And after getting through a few good textbooks you'll be able get through a whole lot more of the glamorous papers.
If you don't know where to start, just find your favorite university's web site and skim syllabi for the classes that interest you. Even better would be to peruse through MIT's Open Courseware, or even registering for classes at a local CC. All of course, if you aren't already headed to an undregraduate degree...
It's amusing, but not surprising: the current paper doesn't need to cite those other papers. It doesn't refer to their research or conclusions. It just counts their citations. The sources for the citation counts, plus sources for the techniques used to analyze the data, should properly be cited.
~Idarubicin
The only conclusion that can be drawn from this "study" is that counting citations is a terrible measure of the relative merits of a paper. It may be OK for comparing average to good papers but obviously fails for evaluating the absolute best discoveries. One simple reason is that because more papers are published today you will get more citations of recent articles - esp since the older ones are established as "fact" and often not cited. If he had done something like normalized for the number of papers being published at the time weighted for how often the offspring were cited, it might have worked a bit better
This type of analysis, while useful for bureaucrats who need simple, if inaccurate metrics, is still dubious. The most cited papers often turn out to be methods papers e.g. how to run gels rather than those with the most import.
I think you'll find that there is more to being a scientist than publishing a good paper.
I perhaps agree, but I have the impression that "publishing good papers" is the key to a scientist having a good career.
Aren't papers the main output of scientists, similar to the tagline "A Mathematician is a device which converts coffee into theorems"?
I really want to believe that "there is more to being a scientist than publishing a good paper" but I'm having a hard time thinking of what that "more" is. Rightly or wrongly, research and academia encourage this with their "publish or perish" attitude.
Tag lost or not installed.
Just referenced, not copied. Many references in a paper are to papers you are comparing with: e.g. disagreeing with. Further, some you just use as background. I suppose you could say extended upon - but usually if you are "extending" a theory it is because there is some problem with the existing theory. Further, once a theory (or, say, an experimental technique) becomes standard, it is no longer referenced and that space is given over to either a review article or a text book.
_ ____
Let me give an example. Suppose someone invents a new technique, say call it "cat splattering spectroscopy" (CSS) and it's useful for looking at widgits. The authors want to get the results out quickly so they submit a Physical Review Letter, max 4 pages, half of which is pictures of widgets. Although the technique may be simple, if it was not discovered until the 21st century, it is likely complicated. Since it is complicated, a page, say, is not sufficient to really explain what CSS entails, but that is all the room they have. This paper is the original, it may be referenced often, but it eventually it will become a poor reference for describing CSS.
Eventually, someone (perhaps the original team, perhaps not) will write a really good paper on CSS and this will be referenced widely, also, eventually replacing the original paper.
Counting citations is a pretty poor way to measure the impact of a paper, but I'm sure these papers are all very good because you can't be cited that many times without having something going for you.
There are many other problems, of course.
________________________________________
a war on terrorism? How can we end a war on a method?