The Nobel prize in physics regularities and tendencies

Jointly published by Akadémiai Kiadó, Budapest Scientometrics, and Kluwer Academic Publishers, Dordrecht Vol. 61, No. 2 (2004) 191 205 The Nobel prize in physics regularities and tendencies ROMUALDAS KARAZIJA, ALINA MOMKAUSKAITĖ Vilnius University Research Institute of Theoretical Physics and Astronomy, Vilnius (Lithuania) Various distributions of the Nobel laureates in physics in the 20th century and their discoveries are considered. It is shown that the time-interval between the discovery and its recognition can be approximately described by a lognormal distribution. The ratio of the numbers of laureates awarded for the experimental and the theoretical discoveries was rather different in various decades; this was determined by some waves of discoveries and in the initial period probably by some subjective factors. The probability to obtain this prize for the theorist is larger than for the experimenter. The main part of the awards was given to the scientists working in the main fields of modern physics: small distances and solid state physics. Some fields of physics such as mathematical physics, relativity, statistical physics were ignored completely. The worrying tendency of an increasing average age of laureates towards their retirement age is indicated. Introduction The Nobel prize is the best known and prestigious scientific prize in the world. There is a large amount of literature on the history of this prize, its founder Alfred Nobel, the procedure of the selection of laureates and their achievements. 1 8 The total number of journal papers with Nobel and Nobel prize in the title included in the data basis ISI Web of Science reached 1500. 9 The bibliography of investigations and data sources till 1978 was given in Ref. 2. Many publications appeared on the occasion of the 100th anniversary of this award. 5 8 In 2002, the material in the Nobel archives related to the first half of the 20th century became accessible to the historians of science. On this ground the investigation of the numerous nominations for the Nobel prize, or Nobel population was performed. 10,11 This population was considered by various aspects: winners and losers, distribution of candidates over different countries, nationalism and internationalism in the nominations, most nominated scientists, predominance of the male candidates, etc. The lists of the Nobel laureates in physics, chemistry and medicine, their distinguished achievements fairly well represent the science of the 20th century. 6,7 Thus, it is an interesting and important task to investigate the distribution of awards and Received June 1, 2004 Address for correspondence: ROMUALDAS KARAZIJA Vilnius University Research Institute of Theoretical Physics and Astronomy A.Goštauto 12, Vilnius 01108, Lithuania E-mail: karazija@itpa.lt 0138 9130/2004/US $ 20.00 Copyright 2004 Akadémiai Kiadó, Budapest All rights reserved

the laureates using the scientometric methodology. It was applied to consider the scientific elite in USA, 12 the relationship between age and productivity of the Nobel prize winners in science, 13 the problem of the attribution of laureates to a certain country and their ranking by birthplace, residence and affilation. 9 We restricted our work to the Prize in physics. Various distributions of laureates or awards will be considered with regard to different fields of physics, time-interval between the discovery and its award, age of the discoverers and laureates, etc. (their nationality is not studied here because it was systematically performed by Braun et al. 9 ). The sources of data The most reliable data on the laureates and the formulations of awards are given on the official Web site of the Nobel Foundation. 1 There also the Nobel lectures, the biographies of laureates and other data, which give the detailed information on the circumstances of discoveries, are presented. This enabled us to establish the field of physics and usually the time of discovery. The systematic material on the Nobel prizes is given in the special encyclopedia. 4 On the celebration of the centennial of this prize the collection of papers 6 was published; it gave a clear view on the history of the Nobel prizes during the 20th century. The article of E. B. Karlson is devoted to the Prize in physics. Each year the main general journals of physics such as Physics Today, Physics World and others publish the exhaustive papers on the newly elected laureates and their main scientific results. The other source was the papers on the history of Nobel prize in physics. 8,11,14 In Refs. 10 and 11 the regularities of the Nobel population in physics and chemistry 1901 1950 were investigated, also some data on the laureates were given. In our work the quantitative statistical investigation is combined with the qualitative analysis of the development of modern physics. Its main trends were discussed in the reviews of the 20th century physics. 15,16 Overall characteristics of the awards Results and their discussion The Nobel prize started exactly with the beginning of the 20th century. Three times it has been given already in the 21st century, however, we will restrict our investigation to a period of the first hundred years, i.e. 1901 2000. It is necessary to note that some known physicists such as E. Rutherford, W. Nernst, I. Joliot-Curie, F. Joliot and O. Hahn obtained the Nobel prize in chemistry for the achievements in the interdisciplinary field of physics and chemistry. Some awards in 192 Scientometrics 61 (2004)

physics were also given for some technical inventions. However, the accurate limits between the adjacent sciences can not be indicated, and we do not take an attempt to include or exclude such cases and will consider only the Prize in physics. During the 20th century 162 scientists have received this honour. The Nobel statute restricts the number of the laureates up to three each year. In the early years this prize was usually given to one scientist. It was not awarded some times during the First and the Second World War (1916, 1940 1942) as well as twice in 1931 and 1934 when the Swedish Academy of Sciences decided not to give the Prize in physics. From the 1960s the average number of laureates during the decade essentially increased and approached to the upper limit 30 (Figure 1). The prize was obtained by one scientist 114 times, it was shared by two scientists in 27 cases and by three scientists in 21 cases. Figure 1. Variation of the number of Nobel laureates in physics during various decades of the 20th century According to the unwritten rule, which was fulfilled traditionally, the scientist was awarded only once during his life even A. Einstein or E. Fermi, who, in the opinion of the historians of science, made 4 5 discoveries worthy of this prize, received it only once. There is one exception the American scientist J. Bardeen was awarded twice for the collective achievements: invention of the transistor and development of the theory of superconductivity. Two awards one in physics and the other in chemistry were received by M. Curie. Scientometrics 61 (2004) 193

On increasing the number of laureates the Nobel prize is sometimes awarded for two achievements from more or less different fields of physics, this is indicated by different formulations. The total number of such formulations was 111 (77 times only one and 17 times two). One formulation stands for the same achievement made together or independently for the experimental registration of phenomenon and its theoretical explanation or for two different, but closely related achievements. The number of the laureates whose achievements were joined by one formulation essentially increased in the second half of the 20th century, it indicated that the discoveries took a more collective character. The Nobel prize in physics only very rarely evokes the discussions that are typical for the Peace or Literature prizes. In the opinion of physicists and the historians of physics 6,8 the selection of the laureates is objective and well-grounded, especially in the second half of the 20th century. Almost all famous creators of modern physics A. Einstein, N. Bohr, E. Fermi, M. Curie, E. Rutherford, M. Planck, W. Heisenberg, E. Schrödinger, L. Landau and many others were the Nobel laureates (as was mentioned above, several of them obtained Prize in chemistry). However, there are some exceptions. L. Boltzmann, A. Poincaré, A. Sommerfeld, G. Gamov were not honoured. A. Sommerfeld received in total 81 nomination before 1950, 11 the most of all physicists, but he was never successful. On the other hand the prize was awarded to the little known Swedish inventor N. Dalén for his development of the automatic light buoy. There are some reasons for these exceptions. The Nobel statute states that the prize should be awarded for the most important discovery or invention, i.e., for a distinguished achievement rather than for a successful life s work. In the early years the Nobel Committee followed rather formally this rule as well as the request that the discovery should be performed recently; the more free interpretation was accepted only later. Secondly, in the face of strong competition among many candidates of high scientific quality the scientist working in the field that has a representative in the Nobel Committee has an advantage (this circumstance will be considered in more detail in the following paragraph). Probably the national sympathies predetermined the award for N. Dalén, but it is necessary to note that this case was a rare exception. There were only two women M. Curie and M. Goeppert-Mayer among the Nobel laureates in physics, which makes less than 1.3%. The similar situation was in chemistry only 3 laureates women, two of them M. Curie and her daughter I. Joliot- Curie worked in the interdisciplinary field of physics and chemistry. However, it would be incorrect to accuse the Nobel Committee of the bias, because the proportion of women among the proposed candidates was even smaller about 1%. 11 Lately some remarkable discoveries, e.g. of the sixth quark or gluons, were made by very large groups of experimenters, and it became difficult to single out one or even 194 Scientometrics 61 (2004)

three main authors. The idea to award all the collaboration was discussed. 17 This is not banned by the statute and even has a precedent of the Peace prize for the Pugwash movement, but such a decision would be still unusual for physicists. Fields of physics How the Nobel awards were distributed over various fields of modern physics? Does it give the objective picture of the development of physics in the 20th century? We have used the main internationally agreed Physics and Astronomy Classification Scheme (PACS) 18 prepared by the American Institute of Physics in collaboration with the International Council on Scientific and Technical Information. In this subsection it is necessary to classify the parts of the annual prize corresponding to the separate formulations. Such a part was named as an award. Sometimes one formulation can include several achievements, e.g., elaboration of the apparatus common to several branches of physics and the investigations carried out by means of them as well as the discoveries in the interdisciplinary area. When the contribution was important in both fields, we attributed this award to both of them equally. The distribution of awards over ten main fields of physics accordingly the sections of PACS is shown in Figure 2. The largest number of awards belongs to the physics of elementary particles (21) and nuclear physics (18). Their distribution in time was different: discoveries of the elementary particles and their laws were mainly awarded in the second half of the 20th century. On the other hand the golden period of the nuclear physics lasted from the end of 1930s till the beginning of 1960s and later the awards in this field became very rare. Indeed, the front-line of physics in the direction of small distances crossed the nuclear physics in the first half of the 20th century, but the elementary particles physics remains the leading area up till now. The general section of physics is not inferior to the indicated two sections. Here 6 awards were obtained for the discoveries of quantum theory and quantum mechanics (mainly in 1918 1933 years). The development of quantum mechanics was closely related with atomic and molecular physics, which achievements were honoured by 13 awards. Thus in total almost half of awards were given for the discoveries in the field which in the first half of 20th century was named as microphysics and conditionally it is often named so now. Indeed, the small distances were the principal direction of the 20th century physics. Scientometrics 61 (2004) 195

Figure 2. Distribution of Nobel awards over various fields of physics accordingly PACS classification. The following abbreviations are used: GEN General; ELEM The physics of elementary particles and fields; NUCL Nuclear physics; ATOM Atomic and molecular physics; CLASS Electromagnetism, optics, acoustics, heat transfer, classical mechanics, and fluid dynamics; PLAS Physics of gases, plasmas, and electric discharges; CON1 Condensed matter: structural, mechanical and thermal properties; CON2 Condensed matter: electronic structure, electrical, magnetic, and optical properties; ASTRO Geophysics, astronomy, and astrophysics The remaining awards (11.5), attributed to the general section of PACS, were given for the new instruments, general experimental methods or precise measurements. This section also includes mathematical physics, relativity, statistical physics and thermodynamics. Neglect of these fields of physics is one of the possible reproaches to the Nobel Committee for Physics. As indicated in Ref. 7, it could be influenced by the absence of the representatives of these fields in the Nobel Committee and the exaggerated caution of some of its members. Even A. Einstein was awarded not for the special or general relativity, but for his services to theoretical physics, and especially for his discovery of the law of the photoelectric effect. The applications of the general relativity to cosmology remained not marked out either. Only the prize in 1993 to R. A. Hulse and J. H. Taylor can be attributed to relativity jointly with astrophysics. Taking in mind the revolution going on in astronomy and cosmology of the 20th century and their close relation with the development of modern physics, the number of awards for this field (7) is relatively small. 196 Scientometrics 61 (2004)

On the other hand, the achievements of the condensed matter physics, its fruitful applications often attracted an attention of the Nobel Committee. In PACS this wide field is divided into two sections: 60 structural, mechanical and thermal properties, 70 electronic structure, electrical, magnetic and optical properties. They received 9 and 14.5 awards correspondingly, thus all in all 23.5. The awards for the investigations in structure of condensed matter were distributed during all the 20th century, whereas in the second, a more refined subfield of condensed matter awards mainly began only from the 1970s. Most awards attributed to the section Electromagnetism, optics, were given for the discoveries in laser physics and its applications. In the 20th century almost all physicists, even of the highest rank, specialized either in the theory or experiment the exceptions were very rare. According to the formulations of awards and scientific biographies of laureates 1,4 they can be divided into the laureates of experimental physics (104), theoretical physics (49) and of both parts (9). Taking into account that the number of scientists experimenters is considerably larger than that of theorists, 15 it may be concluded that the probability to obtain this prize for the theorist is larger than for the experimenter. However, their proportion differed during various decades (Figure 3). On the ground of the analysis of Nobel archives the predominance of the experimenters in the first decade was explained in Refs. 10 and 11 by the absence of the theorists in the initial Nobel Committee for Physics. Figure 3. Ratio of Nobel laureates in experimental and theoretical physics during various decades. The laureates awarded for the achievements of both kinds are not included. The average ratio is shown by a dashed line Scientometrics 61 (2004) 197

However, the large oscillations of the ratio between the experimenters and the theorists in the 7 9th decades must have an objective reason nonmonotonic development of physics. This question is analysed in the following paragraph. Awarded discoveries, age of their authors and the time-interval between the discovery and its recognition Despite the existence of some subjective factors the awarded discoveries make a very representative set of the most important discoveries in physics from the end of the 19th century till the end of the 20th century. Thus the distributions of these discoveries, their authors as well as the time-intervals between the discovery and its award deserve an attention. However, sometimes it is difficult to determine the exact time of the discovery. For a single discovery, as a rule, the date of its publication in a scientific journal is taken. When several achievements in the formulation of award are indicated, often it is possible to establish the main discovery. If the award was given for an integral cycle of results, obtained during some period, the final date should be taken, especially calculating the time-interval between the discovery and its award. There remain some questionable cases (about 7%), which are excluded from our following consideration. The independent different contributions of the winners of the same award are considered as different discoveries. The distribution of the awarded discoveries in time (Figure 4) is rather nonmonotonic some very fruitful periods manifest themselves. The first one in 1891 1914 years was related with the initial discoveries in microphysics: X-rays, electron, radioactivity, quantum idea. The second period 1922 1942 years creation of quantum mechanics, the principles of quantum solid state physics, discovery of the main elementary particles. The third period began after the Second World War and lasted till 1976 it was a very fruitful period for the elementary particles, condensed matter physics and astrophysics, because of the favourable conditions created by an abundant financing from the governmental and private funds. The data for the latest decades of the 20th century can be still complemented by the awards in 21st century. As seen from Figures 3 and 4 generally the experimental discoveries dominated, but at the end of the second period and at the beginning of the third period a series of very important theoretical results mainly in elementary particles and condensed matter physics were made. Most of them were awarded only in the 9th decade, which was the reason of the predominance of awards for theoretical achievements during this decade. 198 Scientometrics 61 (2004)

Figure 4. Distribution of awarded experimental, theoretical and both experimental and theoretical achievements in time Figure 5. Age of scientists at the time of their discoveries Scientometrics 61 (2004) 199

The age of the winners of the Nobel prize in physics at the moment of their main discoveries varies from 22 till 62 years (Figure 5). At the earliest age (at 22) the discoveries were made by O. W. Richardson, G. Marconi and B. D. Josephson, the latest at 62 by C. Cohen-Tannoudji. The average characteristics differs for the theorists and the experimenters: the average age at the moment of the main discovery is Ā exp = 38.3 y, Ā theor = 34 y, and the main deviation equals σ exp = 7.9 y, σ theor = 7.0 y. These results support a widespread opinion that the theorists achieve their highest results earlier than the experimenters. The joined distribution consisting of two distributions shifted in time has a larger width than each of them does (σ = 8.1 y) and a minimum instead of a maximum at the average age Ā = 37.4 y (Figure 5). The coefficient of asymmetry is positive (0.38), this is determined by a longer tail corresponding to the senior ages. The inequality Ā exp > Ā theor also took place for all separate decades of the 20th century (Figure 6) except the first decade, when only two theorists were awarded. Both quantities Ā exp and Ā theor do not show the tendency to decrease or to increase with time. We investigated the distribution of interval D between the discovery and its award as well (Figure 7). This quantity was calculated as a difference of two years do not taking into account a month and a day. Figure 6. Average age of the scientists at the time of their experimental and theoretical discoveries during various decades. (The laureates awarded for the achievements of both kinds are not included.) 200 Scientometrics 61 (2004)

Figure 7. Distribution of the time-intervals between a discovery (experimental, theoretical and both experimental and theoretical) and its award. The approximation of distribution by a lognormal function is shown by a solid line Usually the importance of discovery and its greatest benefit on mankind (according to the formulation of A. Nobel s will) are estimated only after a long time the average time-interval between the discovery and its recognition by the Nobel prize is 15.1 y and its mean deviation from this average value is 11.2 y. Only in four cases the award was given in the next year: it happened with the discoveries of high temperature superconductivity (J. G. Bednorz and K. A. Müller, 1987), W and Z ± intermediate bosons (C. Rubbia, 1984), parity non-conservation (Ch. N. Yang and T. D. Lee, 1957) and for the development of the method of structural analysis by means of X-rays (W. H. Bragg and W. L. Bragg, 1915). The longest records are: the award given to E. Ruska in 1986 after 53 years for the design of the first electron microscope and to S. Chandrasekhar in 1983 forty nine years after the end of his fruitful theoretical studies of the physical processes of importance to the structure and evolution of the stars, and to C. G. Shull in 1994 after 45 years for the development of the neutron diffraction technique. Most of the above indicated very long and very short extremes belong to the experimental physics. The larger portion of the important experimental results were recognized in a period of 1 13 years, but the distribution had a low slowly decreasing tail (Figure 7). The distribution for the theoretical discoveries is less extended in time. The distribution of time-intervals D can be described by a lognormal function 19 Scientometrics 61 (2004) 201

( D ln D) 2 ρ N 1 ( D) = exp 2 2πσ D ln 2σ, ( D 1 ), D ln D where ln D and σ ln D are correspondingly the average value and the mean deviation from it for the distribution of the natural logarithms of time-intervals. The normalization constant N is taken equal to the number of considered discoveries (122). The interval D averaged over a decade (Figure 8) shows a constant tendency to increase starting from 1940s, especially in the last decades of the 20th century. It is a worrying tendency, because the Nobel prize is given only to the living scientists, so the danger increases that the author of the important discovery will wait the recognition in vain. The distribution of the laureates depending on their age at the moment of award (Figure 9) extends within a very large interval: W. L. Bragg won his award at 25 (together with his father) and P. L. Kapitsa only just at the age of 84. The average age of laureates was 52.6 y, the mean deviation from this average 12.1 y, and the coefficient of asymmetry for this distribution is equal to 0.28. Figure 8. Average time-interval between a discovery and its award during various decades 202 Scientometrics 61 (2004)

Figure 9. Distribution of the age of Nobel laureates Figure 10. Average age of Nobel laureates during various decades Scientometrics 61 (2004) 203

The age of laureates averaged over the decade (Figure 10) shows the tendency to increase in the second half of the 20th century, which is related with the increase of the average time of recognition of discovery. Usually the award is obtained not by young or middle-aged scientist as a financial support for his future investigation but by a scientist at an honourable age before he is going to be pensioned off. Conclusions The Nobel prize in physics is highly recognized due to its objective and wellgrounded prize-awarding system (though some subjective factors manifested, especially at the beginning of the 20th century). Thus the distributions of the awards and laureates give a valuable information on the main trends of the modern physics. There were 162 laureates of the Nobel prize in physics during the 20th century. From 1950s this award was often shared by two or three scientists, which testified that the discoveries obtained more collective character. The largest part of the awards was given for the discoveries made in the elementary particles, nuclear, atomic, molecular physics and quantum mechanics; indeed, the small distances were the principal direction of physics in the 20th century. The other most distinguished field was the condensed matter physics very fruitful for its various applications. On the other hand, astrophysics with its many revolutionary discoveries received only 6% of awards and mathematical physics, relativity, cosmology, statistical physics and thermodynamics were ignored at all. The distribution of the awarded discoveries was rather nonmonotonic in time there existed three fruitful periods. Their limits are related with the First and the Second World Wars as well as with the end of the abundant financing of physics. The waves of the experimental and the theoretical discoveries, which mainly determined the different ratio of awards for experimental and theoretical achievements in various decades, may be marked out as well. The average period between the discovery and its award is equal to 15.1 y, only in 4 cases the prize for the achievements of the last year was given. The distribution of this quantity is approximately described by the lognormal law, its slowly decreasing tail mainly belongs to the experimental discoveries. This quantity averaged over a decade tends to increase thus the probability that the author of the discovery will die before the recognition increases. The distribution of laureates with regard to their age at the moment of award extends from 25 till 84 years and the average age tends to increase as well. 204 Scientometrics 61 (2004)

References 1. Nobel Foundation Web Site www.nobel.se 2. J. VLACHÝ, Nobel prizes. A bibliography of scientometric papers and data sources, Scientometrics, 1 (1979) 295 301. 3. N. K. STAEHLE, Alfred Nobel and the Nobel Prizes, The Nobel Foundation, The Swedish Institute, Stokholm, 1993. 4. T. WASSON (Ed.), Nobel Prize Winners. An H. W. Wilson Biographical Dictionary, Vol. 1 2, The H.W.Wilson Company, New York, 1987. 5. B. FELDMAN, The Nobel Prizes: A History of Genius, Controversy and Prestige, Arcade, New York, 2000. 6. A. W. LEVINOVITZ, N. RINGERTZ (Eds), The Nobel Prizes: The First 100 Years, Imperial College Press, London, 2001. 7. R. M. FRIEDMAN, The Politics of Excellence: Behind the Nobel Prize in Science, Henry Holt & Co., New York, 2001. 8. P. RODGERS, Countdown to the Nobel prize, Physics World, 13 (10) (2000) 10 11. 9. T. BRAUN, Z. SZABADI-PERESZTEGI, É. KOVÁCS-NÉMETH, No-bells for ambiguous lists of ranked Nobelists as science indicators of national merit in physics, chemistry and medicine, 1901 2001, Scientometrics, 56 (2003) 3 28. 10. E. CRAWFORD, The Nobel Population: A Census of Nominators and Nominees for the Prizes in Physics and Chemistry 1901 1950, Universal Academy Press, Tokyo, 2001. 11. E. CRAWFORD, Nobel population 1901 1950: Anatomy of a scientific elite, Physics World, 14 (11) (2001) 31 35. 12. H. ZUCKERMAN, Scientific Elite: Nobel Laureates in the United States, The Free Press, 1979. 13. P. E. STEPHAN, S. G. LEVIN, Age and the Nobel-prize revisited, Scientometrics, 28 (1993) 387 399. 14. R. M. FRIEDMAN, Quantum theory and the Nobel prize, Physics World, 15 (8) (2002) 33 38. 15. L. M. BROWN, A. PAIS, B. PIPPARD (Eds), Twentieth Century Physics, Vol. 1 3, Institute of Physics and American Institute of Physics Press, New York, 1995. 16. H. KRAGH, Quantum Generations: A History of Physics in the Twentieth Century, Princeton University Press, 1999. 17. P. RODGERS, Gluons Prize revives discovery debate, Physics World, 8 (9) (1995) 5 6. 18. Physics and Astronomy Classification Scheme 2003, Web site www.aip.org/pacs/. It was used for the scientometric investigation of physics literature for example in: R. TODOROV, Distribution of physics literature, Scientometrics, 7 (1985) 195 209. 19. M. G. KENDALL, A. STUART, The Advanced Theory of Statistics, Charles Griffin, London, 1963. Scientometrics 61 (2004) 205