Saturday 21 July 2007

NFLT metric for revoutionary science - 1987-2006

Editorial

Which are the best nations and institutions for revolutionary science 1987–2006? Analysis using a combined metric of Nobel prizes, Fields medals, Lasker awards and Turing awards (NFLT metric)

Bruce G. Charlton

Medical Hypotheses. 2006; 68: 1191-1194

***

Summary

I have previously suggested that Nobel prizes can be used as a scientometric measurement of ‘revolutionary science’; and that for this purpose it would be better if more Nobel prizes were awarded, especially in three new subjects of mathematics, medicine and computing science which have become major sciences over recent decades. In the following analysis of the last 20 years from 1987 to 2006, I use three prestigious prizes in mathematics (Fields medal), medicine (Lasker award for Clinical Medical Research) and computing science (A.M. Turing award) which are plausible surrogates for Nobel prizes. The combined Nobel–Fields–Lasker–Turing (NFLT) metric is strongly dominated by the USA. However the distribution implies that revolutionary science may be somewhat more broadly distributed than the pure Nobel metric suggests. The UK and France seem to be significant nations in some types of revolutionary science (although the UK has declined substantially as a centre of revolutionary science); and Germany, Switzerland, Japan, Russia, Denmark and Norway also feature. The top world institutions for revolutionary science according to NFLT are MIT, Stanford and Princeton – all in the USA – and the USA has 19 institutions with at least three prize-winners. Second is France, with three institutions having three or more winners; the UK and Norway have one each. The NFLT metric confirms previous observations that many public universities in the Western USA have now become a major focus of revolutionary science; and that Harvard has declined from its previous status as the top world centre of revolutionary science to about seventh-place. This analysis confirms the potential value of increasing the number of Nobel prizes as a means of identifying and monitoring centres of excellence in revolutionary science.


***



Introduction

Revolutionary science is a term coined by Thomas Kuhn in his bookThe structure of Scientific Revolutions (Chicago University Press, 1970) to describe research which changes the fundamental structures of science by making new theories, discoveries or technologies (ie. new ‘paradigms’). But most research is ‘normal science’, comprising checking, trial-and-error improvement and the more gradual and incremental extrapolation of already-existing paradigms.

I have previously suggested that Nobel prizes can be used as a scientometric measurement of ‘revolutionary science’; and that for this purpose it would be better if more Nobel prizes were awarded, especially in three new subjects of mathematics, medicine and computing science which have become major sciences over recent decades [1], [2] and [3]. My three disciplinary suggestions for Nobel expansion are here simply assumed to be valid, and in the following analysis of the last 20 years from 1987–2006, I have used three prestigious prizes in mathematics (Fields Medal), medicine (Lasker Award for Clinical Medical research) and computing science (A.M. Turing award) which are plausible surrogates for Nobel prizes.

The choice of the Fields medal [4] as a near-Nobel equivalent was also made by the well-respected Shanghai Jiao Tong University rankings of the world’s best universities [5]. It is a highly prestigious prize awarded every four years (in batches of up to four winners – making the prize approximately annual) by the International Mathematical Union to a mathematician aged less than 40.

The approximately annual Lasker Award for Clinical Medical Research [6] recognizes from one to three scientists whose work pioneers a major improvement in clinical management or treatment. Unlike the Lasker award for Basic Medical Research, which frequently predicts a Nobel prize in Physiology/Medicine, the Clinical Medical Research (CMR) award does not frequently overlap with the Nobel prize. Only one person in the last twenty years (Barry Marshall) has received both a Lasker award for CMR and also a Nobel prize, and this particular award was removed from the Lasker statistic in the following tabulations.

The A.M. Turing Award is given annually to one or two individuals by the Association for Computing Machinery for contributions of lasting and major importance to the computer field [7].

Having identified the winners of Fields, Lasker and Turing prizes for the past twenty years; I discovered their national and institutional affiliation at the time the prize was awarded – either from the official web pages of the prize-awarding institutions, or by wider internet searching for references to these awards (e.g. Wikipedia entries, press releases, references to the prizes in other publications etc.). Each prize-winner was therefore credited to a single nation and institution. The data from Fields, Lasker and Turing winners were then pooled with the data from Nobel prize-winners and tabulated.

Since the aim of this study was to identify the strongest nations and institutions in revolutionary science, there was a minimum threshold of three Nobel–Fields–Lasker–Turing winners before a nation or institution was included in the tables.

Results

The national Nobel–Fields–Lasker–Turing (NFLT) metric (Table 1) is strongly dominated by the USA, confirming the pattern demonstrated by the previous analysis of Nobel prizes [3]. But inclusion of Fields–Lasker–Turing winners implies that revolutionary science may be more broadly distributed than the pure Nobel metric suggests. The UK and France, in particular, seem to be more significant nations in revolutionary science than suggested by Nobels alone; and other nations are identified as significant which are missed by the purely Nobel prize analysis: Russia, Denmark and Norway.

Table 1.

Number of Nobel–Fields–Lasker–Turing winners by nation 1987–2006
Nation Nobel prizes Other awards Total
USA 126 45 171
UK 9 10 19
France 5 7 12
Germany 9 0 9
Switzerland 7 0 7
Japan 3 1 4
Russia 2 1 3
Denmark 1 2 3
Norway 1 2 3
A minimum of three winners is required for inclusion.

The top world revolutionary science institutions identified by the NFLT metric (Table 2) are MIT, Stanford and Princeton in the USA; and the USA has nineteen institutions with at least three prize-winners. Harvard stays in seventh place for the combined Nobel–Fields–Lasker–Turing metric, which is the same as its Nobel prize-winning rank, tending to confirm my previous observation [3] that Harvard has indeed declined as a centre of revolutionary science – although it remains dominant in ‘normal science’ as measured by metrics such as numbers of publications and citations.

Table 2.

Number of Nobel–Fields–Lasker–Turing winners by institution 1987–2006
US Institution Nobel prizes Other awards Total
MIT 11 2 13
Stanford University 9 1 10
Princeton University 6 4 10
Chicago University 7 1 8
University of California, Berkeley 4 3 7
Columbia University 7 0 7
Harvard University 5 1 6
CalTech 5 0 5
UCSF (University of California San Fransico) 3 2 5
Cornell University 2 2 4
Rockefeller Inst. & Univ. 3 1 4
UCLA 3 1 4
University of Colorado, Boulder 4 0 4
University of Pennsylvania 2 2 4
University of Washington, Seattle 3 1 4
NIH – National Inst. Health 0 3 3
Fred Hutchinson CRC, Seattle 3 0 3
University of California, Santa Barbara 3 0 3
University of California, Irvine 3 0 3

Non-US Institution
University of Cambridge, UK 2 3 5
College de France, Paris 3 0 3
University of Paris-Sud 0 3 3
IHES*, Paris, France 0 3 3
University Oslo, Norway 1 2 3
A minimum of three winners is required for inclusion. IHES = Institut des Hautes Etudes Scientifiques.

In second place to the USA as a home of revolutionary science institutions is France, which has three institutions having three or more Nobel–Fields–Lasker–Turing winners. This particularly reflects French strength in mathematical research, with six Fields medallists in the past 20 years. University of Cambridge (UK) and University of Oslo (Norway) also emerge as significant.

Interpretation

In general, this analysis demonstrates the potential value of increasing the number of Nobel prizes [2], since otherwise the significant strength of France – and its three elite institutions – would be missed. The analysis also confirms the results of the pure-Nobel metric in suggesting that a high level of national performance in revolutionary science is probably a consequence of having elite institutions that win three or more prizes in a 20 year period.

Measured by the NFLT metric; outside of the USA (with its 19 institutions of revolutionary science), only France seems to have succeeded in supporting more than one centre of revolutionary science over the past 20 years. Up until the mid-1980s, the UK was a long-term clear second to the USA in Nobel prizes [3]; but from 1987 to 2006 three of its major prize-winning institutions (i.e. the University of Oxford, the Cambridge Molecular Biology MRC Unit, and Imperial College London) have declined as centres of revolutionary science, and now only the University of Cambridge achieves the three-winner NFLT threshold.

But the most significant result of this analysis is to demonstrate and confirm the massive US domination of revolutionary science [1] and [3], and the lack of any significant national competition for this status except in mathematics (France has six Fields medals to the USA’s eight). This contrasts with the general picture of European, East Asian and Chinese science ‘catching-up’ with the USA in terms of ‘normal science’ production (as measured by numbers of publications and citations [3] and [8]).

Looking into the long term, the lack of international competition in revolutionary science is somewhat worrying, since it means that world scientific progress may increasingly depend upon the US research system. In the US it is probably within-nation research competition between rival institutions that has so far maintained striving and standards in revolutionary science. But if US universities began to compete on the basis of ‘normal science’ instead of revolutionary science – as seems to have happened in the less-diverse, less competitive and more risk-averse Anglo-European research systems [9] – then we might expect to see a decline equivalent to that which has occurred over recent decades in mainland Europe and the UK [3].

The first signs of decline might be seen in previously-successful revolutionary science institutions which, like Harvard or Cambridge (UK), win progressively-fewer major research prizes [3] while maintaining a very high output of highly cited publications [1] and [10]. A more advanced state of decline might be harder to detect, since there would (presumably) continue to be Nobel-, Fields-, Lasker- and Turing-winners even in the absence of actual revolutionary science.

However, at present, the situation 1987–2006 looks healthy and competitive for revolutionary science in the USA, particularly in the elite of MIT, Stanford and Princeton and the recently-emerging Western US institutions [3] exemplified by UCSF with its three Nobels and two Lasker awards (Table 2).

Significance of the NFLT metric

The Nobel–Fields–Lasker–Turing metric only measures the tip of an iceberg of revolutionary science, and my assumption is that each successful example of revolutionary science which has led to a prize, medal or award must (on general theoretical principles [11]) have been supported by a very much larger and more complex system of revolutionary science comprising numerous people and institutions.

So, the NFLT metric will intrinsically register many ‘false negatives’ and systematically under-estimates the scale of revolutionary science. Despite this, the NFLT metric seems to have value, since these prizes apparently have a low false positive rate: impressionistically and anecdotally, the great majority of winners seem thoroughly to ‘deserve’ to win for their research, which has indeed been revolutionary in the sense of changing the direction of science.

It seems unlikely that scientists are frequently, primarily and specifically motivated to do high quality revolutionary science by the prospect of winning a Nobel prize or one of the other high medals and awards – although many would no doubt day-dream about the possibility. Indeed, there is vast variability in the personality types of scientists and their motivations. Rather, Nobels and the like can be seen as providing an after-the-fact identification of some of the clearest and best-validated examples of revolutionary science.

The NFLT metric can therefore be seen as analogous to a ‘top-down’ macroeconomic quantitative variable, such as national taxes and interest rates. Such a variable may have value for monitoring, evaluation and policy; but does not necessarily have a close relationship to individual motivations and behaviors [1]. The NFLT metric is suggestive, but its validity needs to be established by further empirical studies.




Acknowledgement

Thanks to Jonathan Rees and Peter Andras for helpful comments.



References

[1] Charlton BG, Andras P. Evaluating universities using simple scientometric research output metrics: total citation counts per university for a retrospective seven year rolling sample. Minerva [in press].

[2] B.G. Charlton, Why there should be more science Nobel prizes and laureates – and why proportionate credit should be awarded to institutions, Med Hypotheses 68 (2007), pp. 471–473. SummaryPlus | Full Text + Links | PDF (77 K) | View Record in Scopus | Cited By in Scopus

[3] B.G. Charlton, Scientometric identification of elite ‘revolutionary science’ research institutions by analysis of trends in Nobel prizes 1947–2006, Med Hypotheses 68 (2007), pp. 931–934. SummaryPlus | Full Text + Links | PDF (91 K) | View Record in Scopus | Cited By in Scopus

[4] International Mathematical Union. Fields Medal, www.mathunion.org/Prizes/Fields [accessed 19 December 2006].

[5] Shanghai Jiao Tong University: Institute of Higher Education. Academic Ranking of World Universities 2006, http://ed.sjtu.edu.cn/ranking. [accessed 14 December 2006].

[6] Lasker Foundation. Former winners – Clinical Medical research, www.laskerfoundation.org/awards/all_clinical [accessed 19 December 2006].

[7] Association for Computing Machinery. A. M. Turing Award, http://awards.acm.org/homepage.cfm?srt=all&awd=140. [accessed 19 December 2006].

[8] R.D. Shelton and G.M. Holdridge, The EU-US race for leadership of science and technology: Qualitative and quantitative indicators, Scientometrics 60 (2004), pp. 353–363. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus

[9] P. Andras and B.G. Charlton, European science must embrace modernization (Correspondence), Nature 429 (2004), p. 699. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus

[10] Charlton B, Andras P. Oxford University’s research performance (four articles published in Oxford Magazine during 2006). www.hedweb.com/bgcharlton/oxmagarts. [accessed 14 December 2006].

[11] B. Charlton and P. Andras, The modernization imperative, Imprint Academic, Exeter, UK (2003).