the collective construction of scientific memory: the einstein
Hist. Sci., xlvi (2008)
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC
MEMORY: THE EINSTEIN-POINCARÉ CONNECTION AND ITS
Université du Québec à Montréal
“Non ridere, non lugere, neque detestari sed intelligere.”
“I have made a ceaseless effort not to ridicule,
not to bewail nor to scorn human actions,
but to understand them.”
Spinoza, Political treatise (1.4.2)
One of the side effects of commemorations is often to reactivate old debates that
had been forgotten or were lying dormant.1 While this is obvious for political commemorations, it is no less true of scientiﬁc ones, and the UNESCO international year
of physics in 2005, celebrating Einstein’s annus mirabilis of 1905, revived the old
ghost of Poincaré’s contribution to relativity.2 In this context of contested memories,
Olivier Darrigol proposed a kind of “peace treaty” on the basis that “it seems wiser
to acknowledge that Lorentz, Poincaré and Einstein all contributed to the emergence of relativity. That Poincaré and Einstein offered two different versions of this
theory, and that Einstein provided the version that is now judged better”.3 Though
many historians will tend to agree with that generous proposition, I think, on the
contrary, that as an historical assertion (as opposed to a moral one) it is misleading
for two reasons. First it suggests that there was such a thing as a “Poincaré version”
of relativity theory accessible and perceived as such by physicists of the times. As
we shall see in detail by following the reception of Poincaré’s work among physicists, there was never any debate opposing or comparing his “version” of relativity
to Einstein’s or Lorentz’s versions. It is thus a retrospective reconstruction and not
an actor’s category. A second reason for disagreeing with Darrigol’s proposition is
methodological. By suggesting that we give credit to Poincaré for relativity theory,
the historian becomes a judge who presides over the appropriate distribution of credit
in the memorial operation of writing (and rewriting) history, in this case, that of a
central chapter of modern science. As the sociologist of science Robert K. Merton
observed: “... recognition is ﬁnally allocated by those guardians of posthumous
fame, the historians of science. From the most disciplined scholarly works to the
vulgarized and sentimentalized accounts designed for the millions, great attention
is paid to priority of discovery, to the iteration and reiteration of ‘ﬁrsts’. In this way,
many historians of science help maintain the prevailing institutional emphasis on the
importance of priority.”4 So, as Rob Iliffe observed, “historians can even generate
new disputes by attempting to resurrect the case for credit being reallocated to some
0073-2753/08/4601-0075/$10.00 © 2008 Science History Publications Ltd
‘forgotten’ worker”.5 As we shall see, this is exactly what happened in the case of
Poincaré, as scientists, philosophers and historians of science successively entered
the fray to establish “who really did what”.
A striking aspect of Darrigol’s analysis is that its conclusion that both Poincaré
and Einstein (and, of course, Lorentz6) should be credited with relativity is essentially
based on a detailed conceptual analysis of the content of their work. But such an
approach cannot by itself answer the question of what the scientists at the time of
publication did in fact do with these works, and why they rarely if ever considered
that Poincaré should be ‘credited’ with relativity on the same footing as Einstein and
Lorentz. It is well known that there are many contingent reasons that lead a given
community to prefer some contributions to others. Debates surrounding the proper
award of Nobel prizes make this visible: one often sees complaints that scientists
other than the ones chosen had made ‘equivalent’ contributions while others deny
that the contributions were equivalent. One does not have to be a radical relativist to
admit that adjudicating ‘ﬁrsts’ is contingent in many ways and that the historian is
just one more voice among many taking part in the game of distributing “due credit”,
as scientists often say. But the ﬁrst task of the historian — as well as the sociologist
— is less to pass a ﬁnal judgement on who was ‘right’ or ‘ﬁrst’ or ex-aequo than
to inquire into the reasons for that priority debate’s arising in the ﬁrst place and to
explain why actors took the positions they did at a given time.
In other words, a purely internal analysis of Einstein’s and Poincaré’s papers leaves
untouched some crucial historical/sociological questions: Where does this question
of credit to Poincaré for relativity come from? Who raised it ﬁrst? When? In which
context? How did the physicists of the founding period of relativity (1905–11) perceive the supposed links between Poincaré and Einstein?
While trying to provide an answer to these questions, the thesis I will develop in
this paper is that the “Einstein-Poincaré connection” is largely a memorial reconstruction dating from the mid-1950s, revived on the occasion of the celebration in
1994 of the two-hundredth anniversary of the founding of the École Polytechnique,
and culminating with the 150th anniversary of Poincaré’s birth in 2004 and the 2005
World Year of Physics. By contrast, it was not an issue among the scientists involved
in the period 1905–18, that is, before Einstein became celebrated as a public ﬁgure
following the conﬁrmation of his general relativity in November 1919. I thus propose
to solve the ‘mystery’ of the Einstein-Poincaré connection by showing that the physics
community (rightly or wrongly) never hesitated between Einstein and Poincaré when
it dealt with questions related to relativity (that is, “theory of electrons” and “electrodynamics of moving bodies”, to use contemporary terms) and that the ‘mystery’ is an
artefact of projecting backward a particular reading of scientiﬁc papers that does not
correspond to what the actors of the time saw in them. Instead of taking for granted
the legitimacy of the question of ‘priority’ by proposing ‘answers’ to a supposed
‘neglect’ or ‘lack of recognition’, I propose to inquire into the historical conditions
of the emergence of the question itself. As Pierre Bourdieu wrote in another context:
taking the debate as an object is the only way not to take part in it.7
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
2. BIBLIOMETRICS: A METHOD FOR THE GLOBAL ANALYSIS OF A DISCIPLINE
By using a bibliometric methodology adapted to the analysis of a whole community,
as opposed to the biographical and conceptual analysis of a single or a few scientists,
we can shed new light on the way Einstein, Poincaré and Lorentz were used (and
connected) by the participants in the physics community of publishing researchers.
My analysis is based on the hypothesis that we can get a good indicator of the
collective perception of the links between published works by looking at the cocitations in the published papers over a given period.8 It goes much further than the
simple counting of the number of papers, which is a useful but limited indicator of
foci of interest.9 We thus get a global access to the connections made ‘in real time’
so to speak (and not with hindsight) among existing contributions. In other words,
co-citation analysis gives us a view of the connections made by the actors themselves
as opposed to those created by historians and philosophers of science through their
conceptual analysis. The advantage of that approach over an analysis of the ‘internal
logic’ of the papers is that it registers the use (or non-use) of given papers without
trying to evaluate their ‘originality’ or ‘real meaning’, or establishing their ‘common’
content or who was really the ‘ﬁrst’ to propose an ‘idea’. The task of evaluation
should be left to the actors involved as it is highly probable that what historians
now see as related or even identical was not seen that way by the actors involved at
the time. As Darrigol rightly points out, “active, ﬁltering reading of scientiﬁc texts
is very common”.10 In fact, it is unavoidable as reading a paper is always done in
the context of particular resources and interests. Andrew Warwick also insisted that
when it was read in Cambridge, Einstein’s paper on relativity “took on new meanings
that can only be understood in terms of the local training and research practices”.11
Of course, what is true for Einstein is also true for Poincaré and such an historicist
approach cannot provide an ‘assessment’ of the respective contributions of scientists
to relativity theory,12 as it limits itself to trying to understand what in fact the actors
did at the time.
It is now well known, as I have already noted, that as early as 1906, the theory
was very often referred to as the “Lorentz-Einstein” theory, thus explicitly connecting Lorentz to Einstein. If the bibliometric method is appropriate, it should bear this
out by showing a larger frequency of co-citation between Einstein and Lorentz than
between Einstein and Poincaré. Conversely, a very low level of co-citation between
Einstein and Poincaré would suggest that actors, for whatever reasons, did not perceive these authors as closely related. As we will see, the results obtained for the
period 1905–44 are consistent with this view. Moreover, extended over the period
1945–79, the latter year being the centennial of Einstein’s birth, the results strongly
suggest that the “Einstein-Poincaré connection” developed only in the second half
of the twentieth century, particularly in historical and philosophical papers assessing
the ‘birth’ of relativity.
It is worth noting that despite frequent use of the term ‘scientiﬁc community’ and
‘discipline’ in many historical papers, most of them adopt an individualistic approach
and concentrate on a few scientists, using the usual resources: analysing the content of
their papers (and sometimes the correspondence), and taking it for granted that they
probably represent the trend of the whole community and that empathy and immersion make it possible to read the papers as scientists did at the time. The method I use
here provides a way to tackle the community as a whole without looking at particular
scientists and into their individual choices. In a Durkheimian fashion I am interested
in the collective behaviour of scientists and not their individual reasons, which vary,
for citing or not citing their peers. Of course the two approaches are complementary
and can also be combined; for once a pattern has been established at the level of a
discipline or a specialty, one can then try to explain the particular pattern of citations
and co-citations by invoking cognitive and sociological reasons.
This collective analysis eschews any psychologistic explanation, which is based on
implicit moral evaluations of the behaviour of agents, a tendency difﬁcult to avoid at
the biographical level. Hence it is frequent to read that it is “puzzling” that Poincaré
was not often cited by Lorentz or Minkowski or others, given his contribution, or that
it is “surprising” that the collection of basic papers on relativity published in 1913
did not contain Poincaré’s 1906 paper.13 Not only are these seemingly innocuous
formulations rarely followed by any sustained analysis providing an explanation of
the ‘puzzle’ but, more importantly, they take for granted what from an historicist point
of view should be at stake, namely, that they had to cite him.14 In fact, an historicist
approach looks at the problem the other way around: given the historical fact that
most physicists usually did not refer to Poincaré in the papers on “relativity”, this
may be simply because they did not see in it what we (or at least some of us) see in it.
And even if the individualistic approach can indeed ﬁnd good reasons why someone
in particular failed to refer to Poincaré, the question remains of the global behaviour
of the community; and, short of our accepting a conspiracy theory,15 this collective
behaviour requires a socio-cognitive analysis that transcends particular cases.
After having brieﬂy presented the database used in this paper, and analysed the
various results obtained from it, I will discuss the evolving contexts in which the
question of the contribution of Poincaré to relativity has been raised at different
moments since the end of the First World War. In so doing, I will suggest cognitive
and sociological elements contributing to an explanation of the reading (and reception) of Poincaré’s work on the theory of the electron by the physics community of
the time, in light of the results obtained. I think that the existing historical literature
already contains many of the elements of the explanation I propose, but that these
clues have simply not been put together for lack of an appropriate theoretical and
methodological framework. By concentrating on the intrinsic conceptual structures
of the published papers, historians are directed toward a rationalist reconstruction
of the theory that is disconnected from the collective behaviour of the scientists
who used the work of their colleagues. While comparing Einstein’s and Poincaré’s
papers may show that we today as historians see how much they are related, that
does not answer the historical question: why did most physicists at the time not see
what historians now see? Only a purely historicist reading of the history of relativity
based on actors’ practices can get around the spontaneous tendency of attributing
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
TABLE 1. Papers and references in selected source journals (1900–44).
1 295 014
2 237 583
Source: Thomson Scientiﬁc Century of science database.
credit and ‘evaluating’ the contribution of each actor. An historicist approach also
helps make visible the sociological basis of the very question of attributing ‘proper’
credit. In other words: a purely conceptual analysis of the contributions of Poincaré
and Einstein can only lead to questions about the proper attribution of “who really
did what ﬁrst”, even when this question was never at issue in the community itself but
appeared only ‘after the battle’ at moments of celebrations and commemorations.
3. THE DATABASE
The bibliometric analysis is based on Thomson Scientiﬁc’s Science Citation Index
(SCI) for the period 1900–79. The period 1900–44, covered in the Century of science database, comprises papers and references published in 266 journals.16 I have
grouped them by disciplines and assigned 27 journals to physics largely deﬁned
(including astrophysics and astronomy), 14 to mathematics, 34 to chemistry, and 10
to “general” since they cover academy journals that are multidisciplinary and cover
physics, mathematics as well as other domains.17 As Table 1 shows, the analysis of
the citations for the period 1900–44 is based on 308,435 papers containing more
than two million references (2,237,583). In journals like Nature, many articles have
no author names although they may contain references. There are also many papers
without references, and the average number of references per paper containing at
least one reference varies with discipline from about 8 in mathematics to about 13
It is often said that Einstein was not particularly prone to referring to earlier papers.
Einstein himself admitted as much and our database makes it possible to calculate
the average number of references per paper and thus compare referencing practices
of different authors. For the period 1900–14, the average of the two main German
physics journals (Annalen der Physik and Zeitschrift für Physik) is 8.8 whereas Einstein refers only to 2.7 articles per paper compared to 6.2 for Max Planck and 5.5
for Wilhelm Wien in the same period. The number of references per paper tends to
rise with time, as more papers get published thus creating a larger pool of potential
references. Of course, this number also depends on the discipline and specialty, as
FIG. 1. Evolution of citations to Einstein and Poincaré (1900–79).
4. CITATIONS TO EINSTEIN’S AND POINCARÉ’S “RELATIVITY” PAPERS
Before concentrating on the citations to the speciﬁc contributions to relativity in 1905
and 1906 by Einstein and Poincaré, we will look at the general evolution of total
citations to all of their published works over the period 1905–79. The centenary of
Einstein’s birth provides a convenient end point as it generates a peak of interest but
no new trend for the analysis.
Figure 1 clearly shows that the ‘take-off’, so to speak, of citations to Einstein dates
from the end of the Second World War and continues unabated as he becomes an
historical ﬁgure following his death in 1955. Part of the growth is due to the addition
of social science and humanities journals to the Thomson Scientiﬁc database in the
mid-1960s, but even with this discontinuity in the source data, it remains obvious
that the interest in Einstein’s work, among scientists as well as among humanists,
has always been much higher than in Poincaré’s, despite the fact that both have
been considered scientiﬁc ‘geniuses’. A very interesting characteristic of Poincaré’s
citation curve is that it clearly starts to rise only in the mid-1950s, signiﬁcantly later
than is the case for Einstein. As we will see in more detail below (Section 10), this
development can be attributed to the publication and impact of the second volume
of Edmund Whittaker’s History of the theories of aether and electricity, where he
strongly argued against Einstein’s paternity of relativity, which he rechristened the
Whereas total citations give us for the period a general view of the increase of
interest in Einstein’s and Poincaré’s work, our thesis will be more convincing if it still
holds for the sub-set of citations to the speciﬁc papers related to relativity: Einstein’s
1905 paper on the electrodynamics of moving bodies and Poincaré’s two papers on
“La dynamique de l’électron” (the brief note in the Comptes rendus hebdomadaires
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
FIG. 2. Citations to “relativity” papers of Einstein (1905) and Poincaré (1905, 1906), in the period
FIG. 3. Citations to “relativity” papers of Einstein (1905) and Poincaré (1905, 1906), in the period
de l’Académie des Sciences of June 1905 and the much longer paper in the Rendiconti
del Circolo Mathematico di Palermo of 1906). I combined the citations to the latter
two papers to make the comparison with Einstein’s paper. The results are shown in
Figure 2 for the period 1905–39 and in Figure 3 for the period 1945–79.
A comparison of these two ﬁgures makes it clear that in scientiﬁc journals before
1945 the numbers of citations to Poincaré’s two papers are much lower than those
to Einstein’s 1905 paper, and that they rise signiﬁcantly only after the mid-1950s.
This feature (also observed in Figure 1) again points to the probable historiographical
impact of Whittaker’s book. With 35 citations over the period 1905–11, Einstein’s
paper was cited ﬁve times more frequently than Poincaré’s contribution on the
subject, which garnered only seven citations in the scientiﬁc journals covered in the
database.19 The ratio goes down to 3:1 for the period 1905–53 (62 against 21), and
during the period 1954–79 it goes down again, to 2:1 (229 against 103).
5. CO-CITATIONS TO EINSTEIN, POINCARÉ AND LORENTZ
A last measure of these trends, and a more signiﬁcant one, is provided by co-citations. Whereas citations give us a measure of the total credit (or symbolic capital20)
given to scientists by their peers, co-citations provide a measure of the connections
between authors, as perceived by the scientists making the citations.21 Moreover, citations (and thus co-citations also) being chosen by scientists, they provide a window
on the extent to which members active in the ﬁeld choose to create links between
different contributions, thus bringing attention to them.22 And despite its limitations,
and the fact that some citations remain implicit and thus lower the ‘exact’ count, the
general trend of citations (and co-citations) provides a good indicator, for one can
hardly imagine that, at the level of a ﬁeld, a ‘famous’ contribution would never be
cited by anybody, either positively or negatively. It is also not realistic to imagine a
paper implicitly discussed but never explicitly cited.23 Also, after a variable period
of years, the level of citations for a given contribution goes down as it becomes
well known and taken for granted, thus becoming implicit and not registered by an
explicit citation. More contingent reasons can also affect citations, as was the case for
Einstein in the Nazi era where it was not prudent to cite him in a paper even when it
was about relativity.24 But again, as long as we analyse trends, these caveats do not
affect the general conclusions.
Figure 4 shows the evolution of the number of co-citations to Einstein’s and
FIG. 4. Co-citations to Einstein and Poincaré and to Einstein and Lorentz (1906–39).
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
FIG. 5. Co-citations to Einstein and Poincaré and to Einstein and Lorentz (1940–79).
Poincaré’s work, all papers included. The idea behind this indicator is that if the
authors are generally seen as working on closely related topics, then the amount of
co-citations should be signiﬁcantly higher for them than for scientists working on less
related topics. For comparisons, I also include the number of co-citations to Einstein
and Lorentz, since we expect the latter to be much larger than the former. And once
this connection is created, it is to be expected that there will be a proximity effect at
work: scientists often co-cited for a given kind of work will tend to be co-cited for
other similar works. Also, as the sociologist Robert K. Merton has shown, those who
already have credit will tend to get more than those who have less, all other things
being equal.25 This “Matthew effect” is linked to the discipline since the credit of a
scientist in a given ﬁeld (in this case physics) is not directly applicable in another
ﬁeld (mathematics for example). This sociological law affecting the distribution of
credit among members of a ﬁeld certainly played a role in the credit accorded by
physicists to Poincaré’s work on “relativity”.
Like the pattern observed in Figures 1, 2 and 3, it is clear that the amount of cocitations to Einstein and Poincaré (shown in Figures 4 and 5) rises signiﬁcantly only
after the mid-1950s. Before 1945 it is much more frequent to encounter “EinsteinLorentz” co-citations than “Einstein-Poincaré” co-citations. For the period 1900–11
the ratio is more than 3:1 (27 co-citations to Einstein-Lorentz as opposed to 8 for
Einstein-Poincaré, the ﬁrst of which being by Einstein himself26). Interestingly, the
two co-citation levels reach the same order of magnitude only after 1954, with respectively 39 and 30 between 1954 and 1964 and 150 and 141 between 1965 and 1979. It
is thus clear that only after the mid-1950s did scientists, historians and philosophers
of science start to create strong links between Einstein and Poincaré.
As should be expected, for the period 1900–44, more than half of the total of the
29 co-citations to Einstein and Poincaré are concentrated in the three main physics
journal of the time: Annalen der Physik (with 10), Zeitschrift für Physik (6) and
Physikalische Zeitschrift (2), the rest being scattered in eleven different journals
including two mathematical journals (Acta mathematica and Mathematische Annalen)
and three French journals (Compte rendus de l’Académie des Sciences, Journal de
physique et de radium, and Annales de chimie et de physique). Among these cocitations, only three are to the speciﬁc papers on “relativity” in the period 1905–11.
The low level of citation to Poincaré’s dynamics of the electron in the formative
period of relativity is also due in good part to the fact that his paper was not only in
French but published in an Italian journal of mathematics. These three details should
not be discounted as insigniﬁcant and it is worth reminding ourselves that ideas do
not circulate ‘in the air’ but in journals, and some are less visible (and accessible)
than others.27 Physicists had no particular reason regularly to turn to the Rendiconti
when scanning the literature. Citation data for the period 1900–44 conﬁrm that this
journal was used essentially by mathematicians, and even in this ﬁeld it was less
visible than the major mathematical journals of the times. Hence, the Rendiconti
gets about 800 citations over that period and only about 12% of these come from
physics journals. By comparison, Acta mathematica (where Poincaré sent most of
his major papers) gets about 2000 citations (excluding self-citations) with only 10%
from physics journals, while Mathematiche Annalen gets about 6000 citations with
again 12% from physics journals. Journals were thus essentially disciplinary, as is
also suggested by co-citations maps (see below, Figures 8 and 9), and it is probable
that they were not consulted regularly by scientists outside the discipline.
For the period 1945–79, the pattern of co-citations is very different. As shown in
Table 2, the co-citations are more frequently found in journals of history and philosophy of science and in pedagogical journals such as American journal of physics.
I will analyse this period in more detail in Section 10.
Summarizing all the above data, we can say that before the Second World War,
Einstein’s work was read by practising scientists as closely related to that of Lorentz
and very rarely related to that of Poincaré. Nothing suggests that among the members
of the scientiﬁc community of that period, the question of Poincaré’s contribution
to the creation of “relativity theory” was ever an issue or that the label “LorentzPoincaré” has ever been used to describe the special theory of relativity, whereas the
expression “Lorentz-Einstein” was frequent in the literature. The most revealing part
of the data are those for the period 1955–79 where the connection between Einstein
and Poincaré becomes as important as the older one between Einstein and Lorentz.
But before turning to the question of the emergence of that new pattern, let us look
more closely at the rise of Einstein in the ﬁeld of physics and at Lorentz’s reading
of the contributions of Einstein and Poincaré.
6. THE RISE OF ALBERT EINSTEIN
As we can see in Figure 6, the annual count of citations to Lorentz and Poincaré is
of the same order of magnitude for both, something that could be expected given
their recognized status as leader in their respective ﬁelds. They are also among the
most cited (Poincaré is fourth over the 1900–11 period in mathematics and Lorentz
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
TABLE 2. Co-citations to Einstein and Poincaré by source journal (1945–79) (for n>3).
American journal of physics
The British journal for the philosophy of science
Philosophy of science
Reviews of modern physics
Reports on progress in physics
Source: Thomson Scientiﬁc Web of science.
FIG. 6. Evolution of citations to Poincaré and Lorentz by citing year (1900–11).
twelfth in physics, the average per year being about 2 per author in physics and 1.5
in mathematics). By comparison, Einstein’s visibility in the physics community
rises sharply after 1905 to attain the level of Lorentz and Poincaré by 1910 (Figure
7). Also, Lorentz and Einstein being essentially theoretical physicists, most of their
citations come from physics journals.28 Poincaré is a more universal ﬁgure who, in
addition to being a mathematician of genius, also covers physics and philosophy, and
the distribution of his citations by the discipline of the citing journals varies accordingly. As could be expected, Poincaré is ﬁrst cited by mathematical journals (48% of
the citations) but still gets about 37% of his citations from physics journals (Table 3).
The works cited are also different: physicists ﬁrst cite his books on electromagnetism
and optics, whereas mathematicians cite his many papers in Acta mathematica and
the Comptes rendus de l’Académie des Sciences.
A more striking measure of the centrality of the three actors in their respective ﬁelds
is provided by calculating the degree of centrality of each of them in the co-citation
network.29 Figure 8 shows the links between all co-cited scientists (above a threshold
TABLE 3. Citation by discipline of citing journals (1900–11).
Source: Thomson Scientiﬁc Century of science database.
of 4 co-citations) in the physics journals of the database for the period 1905–11. It
clearly shows the strong links uniting Einstein, Lorentz, Planck and Abraham, while
Poincaré is absent from the map. Figure 9 shows the equivalent map for mathematics
over the period 1900–11, in which Poincaré is central and strongly related to F. Klein,
H. A. Schwarz and E. Picard. In terms of centrality, Poincaré ranks fourth after D.
Hilbert, Klein and Picard, all well-known mathematicians of the times.
Interestingly, the rise of Einstein is directly visible in the evolution of his degree of
centrality: while he was absent in the period 1900–05, he jumps to the tenth position
in the 1905–11 map of physics co-citations, while Lorentz is eighth (behind J. J.
Thomson, P. Drude, J. Stark, P. Lenard, M. Planck, E. Warburg, and M. Abraham)
and W. Voigt is ninth. A calculation of the degree of centrality for the next period
(1912–18) shows Einstein moving up again to fourth position while Lorentz moves
back to twelfth and Drude to ﬁfteenth. Their declining centrality compared to Einstein
suggests an important restructuring of the ﬁeld after special relativity made models
of the electron obsolete, while general relativity was being developed and quantum
theory had become a lively topic, both of these having Einstein as a central actor.30
Finally, as the two maps suggest, there are no signiﬁcant links between the ﬁeld of
FIG. 7. Citations to Einstein (1900–11).
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
FIG. 8. Co-citation network of authors in physics journals (1905–11). Only links with 4 co-citations or
more are shown.
mathematics and that of physics.
7. LORENTZ’S READING OF POINCARÉ’S LA DYNAMIQUE DE L’ÉLECTRON
As the central ﬁgure in the research domain of the “theory of electrons”, H. A. Lorentz
is probably the best lens through which one can access the reading most physicists
had made of Poincaré’s contribution to that ﬁeld.31 Both men were well known at the
turn of the century and they respected each other. They also exchanged letters, and
Poincaré studied in critical detail Lorentz’s theory of electrons. Given that Lorentz
was a correspondent of Poincaré, and nominated him in 1910 for the Nobel Prize for
Physics, it is interesting to look at his comments on Poincaré’s Rendiconti paper. His
vision was probably typical of his fellow physicists and contributes towards explaining
the lack of signiﬁcant co-citations linking Poincaré directly to relativity.
In a brief letter dated 8 March 1906, he wrote to Poincaré to acknowledge the
receipt of his “important memoir on the dynamics of the electron”. In a crucial passage (rarely quoted) Lorentz writes: “Needless to say, I studied it with the greatest
interest and I have been very glad to see my conclusions conﬁrmed by your considerations.”32 What this sentence suggests is that Poincaré is perceived as a great
FIG. 9. Co-citation network of authors in mathematics journals (1900–11). Only links with 4 co-citations
or more are shown.
mind clarifying aspects of already existing theories but not himself being a central
and ‘full-time’ actor in their creation in the physics community. The fact, mentioned
above, that physicists cite his textbooks also conﬁrms this view, as these books are
critical syntheses and not the standard participation in the community via the publication of papers in physics journals. And it is not farfetched to suggest that stylistic
traits of Poincaré’s paper also contributed to this reading. In fact, Lorentz’s sentence
quoted above is a paraphrase of Poincaré’s assessment of his own contribution in the
introduction of his Rendiconti paper, where he writes: “the results I have obtained
agree with those of Mr Lorentz in all the important points; I was led to modify and to
complete them only in a few points of detail; the differences, which are of secondary
importance, will be seen later on.”33 Even taking into account the rhetorical tradition
of understating one’s own contributions, it still suggests a simple conﬁrmation of
the basic work of Lorentz.
In matter of credit attribution, there is always, in addition to a national bias clearly
visible for example in the nominations for Nobel prizes,34 a disciplinary bias. As
members of a discipline, scientists tend to attribute more credit to their peers than to
scientists from other disciplines seen as intruding ‘part time’ into their ﬁelds and as
being somehow external to their community.35 From their point of view, it is probable
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
that Poincaré’s works in physics were indeed perceived as a “critique of physical
theories”, as Poincaré himself called many of his contributions.36
These facts are more than anecdotal when one takes into account the sociological
aspects of being part of a community as a regular and active contributor to a ﬁeld.
Those who devote themselves completely to the task as physicists (experimental and
theoretical) will tend to see the sporadic critical comments of a mathematician as
useful but not central, and will tend to credit the contributions of those who did the
complete physical development, and not only suggested the general idea and corrected the mathematical deﬁciencies of the physicists. We can see a conﬁrmation of
this tendency in the fact that what attracted Lorentz’s attention in Poincaré’s paper
was what became known as the “Poincaré pressure”. In the series of lectures on the
theory of electrons, given at the University of Columbia at the end of March 1906
(just two weeks after he wrote the letter to Poincaré quoted above) and which were
published in 1909, he refers twice to Poincaré’s 1906 paper. In both cases, it is to
mention the pressure needed to keep the electron in equilibrium against the electrostatic force which would, alone, make it explode.37 By contrast, in the preface of
that book, he noted that he could not devote as much space as he would have liked to
“Einstein’s principle of relativity” which has not “received an adequate treatment”.38
Nonetheless, paragraph 189 of the ﬁrst edition is explicitly devoted to “Einstein’s
theory”. In the second edition of 1915, a long note is added to explain in more detail
this theory, but without making any connection to Poincaré’s paper. Even in his later
book Problems of modern physics, a course of lectures delivered at Caltech in 1922,
he still refers to Poincaré only in relation to “Poincaré’s model of the electron” and
its pressure.39 Finally, in 1927, in a talk at Mount Wilson Observatory he repeated
that “the theory of relativity is really solely Einstein’s work. And there can be no
doubt that he would have conceived it even if the work of all his predecessors in the
theory of this ﬁeld had not been done at all. His work in this respect is independent
of the previous theories”.40
Though Lorentz got a copy of the paper directly from Poincaré, there are indications
that his Rendiconti paper may not have been so easy to ﬁnd for the average physicist.
As late as 5 July 1909, a long-time friend of Poincaré, the Swedish mathematician
Gösta Mittag-Lefﬂer, wrote to Poincaré: “You undoubtedly know the pamphlet by
Minkowski ‘Raum und Zeit’ published after his death as well as the ideas of Einstein41
and Lorentz on the same question. Now M. Fredholm tells me that you have touched
upon similar ideas before the others, while expressing yourself in a less philosophical, more mathematical manner.”42 He asked him if he could write a paper on the
subject, which he would publish in his own mathematical journal (Acta mathematica).
Written in a “language comprehensible by simple geometers”, it would render “a
great service to everyone”. As the editors of the Poincaré’s correspondence noted,
the two mathematicians (Fredholm and Mittag-Lefﬂer) were probably preparing
themselves to propose Poincaré for the 1910 Nobel Prize for Physics and were thus
revising his contributions to physics. Now, if a close friend like Mittag-Lefﬂer could
ignore the 1905 and 1906 papers three years after their publication, the chances are
that physicists were even less likely to know it, for it appeared, as we have seen, in
an Italian mathematical journal and in French.
The strategy adopted to get a Nobel prize for Poincaré, under the leadership of
Mittag-Lefﬂer, also sheds light on what the actors of the time perceived as his most
original contributions to physics. They all promoted his contributions to the “differential equations of mathematical physics”, but some nominators did also refer to
his Rendiconti paper on the dynamics of the electron in their list of important papers.
In their report to the Nobel Committee, the mathematicians Darboux, Fredholm and
Appel wrote that Poincaré contributed “very important results” to the “very timely
and very delicate” theory of electrons by studying “the forces acting between the
diverse parts of an electron”, and that he found that the Lorentz contraction was made
very plausible once considered in the light of the principle of relativity. They also
highlighted his results concerning the ﬁnite speed of gravity that should propagate
at the speed of light, and recalled his criticism of Lorentz’s theory concerning the
applicability of the principle of reaction, a result that, they added, lies at the basis
of Max Abraham’s electromagnetic conception of the electron.43 The Italian mathematician Vito Volterra also noted in his letter to the Nobel Committee that Poincaré
contributed to the recent discussions surrounding the principle of relativity by looking
at the question from a very high level, showing that through the principle of least
action the postulate could be proved if one accepted the electromagnetic constitution
of the electron plus the existence of a non-electromagnetic constant pressure which
explains the Lorentz contraction.44 Here again we ﬁnd reference to the structure
of the electron and the pressure needed to keep it stable. But despite this vigorous
campaign, Poincaré was never awarded the Nobel Prize for Physics.45 At about the
same time, in 1912, the physicists Wilhelm Wien and Clemens Schaefer proposed
that the prize be given jointly to Lorentz and Einstein for their work on relativity, a
further indication that the dominant perception of physicists was that the theory of
relativity originated essentially with these two.46
Following Poincaré’s death in July 1912, a series of eulogies were published and
none of them raised the question of the lack of proper recognition for his contributions to electron theory or suggested that they were equivalent to Einstein’s relativity
theory, which was, by then, well known in physics. The question of ‘priority’ was
not yet an issue, not even for the French physicist Paul Langevin. In a very long
analysis of “Poincaré le physicien”, published in 1913, Langevin, who knew him
well and was himself the ﬁrst French scientist to promote Einstein’s work on relativity, provided a history of Lorentz theory and the principle of relativity.47 He noted
that, in addition to writing numerous works devoted “to the exposition, discussion
and comparison” of the diverse electromagnetic theories (Maxwell, Larmor, Hertz,
Lorentz, etc.), Poincaré made a “decisive contribution” by introducing the notion of
electromagnetic momentum and also called attention to his invention of the “pressure” needed to keep the electron stable.48 Presenting the contribution of Lorentz to
the theory of electrons, he added that the Dutch physicist did not draw the ultimate
consequences of his own work and kept the idea of absolute time. In the next paragraph
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
he mentioned that Einstein made things “clearer” by pointing at the new notions of
space and time and by giving to the principle of relativity its full generality, also
giving the “deﬁnite form to the Lorentz group”. Though it is impossible to know
exactly what he meant by “clearer” it is very likely related to the fact that Einstein
derived the Lorentz transformations from two independent principles, whereas all
others (Lorentz as well as Poincaré) were using them as postulates without knowing
“where they come from”, so to speak. The following paragraph of Langevin’s “continuist” history introduced Poincaré who “arrived simultaneously at the same result
using a different road, his attention having been attracted ﬁrst by the imperfect form
of the transformations as given by Lorentz”. So, even Langevin did not yet talk of
Poincaré as a “precursor” and he selected the same contributions (momentum and
pressure) as did all of physicists who referred to his 1906 paper. He also pointed to
the fact that Poincaré was perfecting Lorentz’s work as he was attracted to it by “the
imperfect form” of the Lorentz transformations. Finally, mentioning the private conversations he had with him during their trip to the Saint-Louis International Congress
on Arts and Science in 1904, he noted that Poincaré “followed with passion all the
phases of the revolution that was transforming our most fundamental conceptions”,
but added that he saw with some concern the old ediﬁce of Newtonian dynamics
being shaken. By using the word “followed” Langevin also implicitly suggests that
Poincaré’s contributions, despite their importance and profundity, were always done
from a certain overarching critical distance.
While preparing a volume dedicated to the memory of Poincaré, Mittag-Lefﬂer, the
long-time chief promoter of Poincaré’s genius, wrote to Einstein in mid-December
1919. Describing the plan of the volume, he mentioned the contributions of Lorentz,
Wien and Planck, and asked him to contribute a paper “lacking there, on Poincaré’s
ideas concerning the important theory created mainly by you, the theory on the
relationship between space, matter, and time”.49 Though Einstein did not contribute
to the volume,50 it is interesting to note that while Lorentz’s contribution discussed
Poincaré’s two papers on the dynamics of the electron, Mittag-Lefﬂer still considered
his volume needed another paper discussing the “relationship between space, matter,
and time”, as if the mathematician considered these aspects to be distinct from the
theory of electron covered in Poincaré’s papers.
In his contribution to the volume of Acta mathematica dedicated to Poincaré
and ﬁnally published in 1921, W. Wien, who a few years before had recommended
Einstein and Lorentz for the Nobel prize, surveyed Poincaré’s contributions in
physics, and mentioned (in addition to his works in statistical mechanics, the solar
system and telegraphic transmissions) his notions of electromagnetic momentum
and the internal forces at work on the electron (Poincaré pressure). He noted that
on the topic of “the modern theory of relativity he contributed important results”,
and “foresaw [vorausgesehen] the general mathematical relations occurring in this
theory”, such as the four-dimensional formulation of the Lorentz transformations.51
As for Lorentz, he brieﬂy recalled the train of thought that led to the principle of
relativity, noting that his own work was greatly “stimulated by the kind interest
that Poincaré accorded” to it. Sensitive to Poincaré’s comments that the contraction
hypothesis was somewhat arbitrary, he was pushed toward a “more general theory”
which could explain Michelson’s results as well as all others trying to detect relative
motion of the second order in v/c. Coming to the transformation equations, he noted
that it was these considerations published in 1904 that brought Poincaré to write his
memoir on the dynamics of the electron “in which he attached [Lorentz’s] name to the
transformations just mentioned”. But, he added, they were not perfect, and for some
physical variables, he did not ﬁnd the most convenient form of the transformation,
something which “Poincaré did, followed by Einstein and Minkowski”. Lorentz also
mentioned that Poincaré was the ﬁrst to use the term ‘postulate of relativity’. Moreover
he “corrected the imperfections of [Lorentz’s] work without ever reproaching [him]
for them”. Like many physicists, Lorentz also mentioned Poincaré’s introduction of
the non-electromagnetic pressure and concluded his long analysis by recalling that
he introduced new notions concerning rotations in four dimensions. He mentioned
these ideas, he said, “because they recall methods Minkowski and others would later
use to facilitate the mathematical operations used in relativity theory”.52
What all these examples clearly show is that Poincaré’s work on the theory of
electrons was not ‘neglected’, nor did it go unrecognized by his contemporaries. It
has simply been read in a way that makes it a useful but not a ‘revolutionary’ or even
central contribution to be set apart from all the other contributions made by physicists,
except for the ideas of electromagnetic momentum and the non-electromagnetic pressure keeping the electron in equilibrium, which were cited as such by many physicists
of the time. Michel Paty probably came closest to the contemporary judgements on
Poincaré’s work on the physics of electrons when he wrote that Poincaré’s program
was “to complete the theory of Lorentz while simplifying and correcting it on minor
points, to make it more rigorous”.53 But mathematicians’ view of his work was different from that of physicists, as we have already seen in the context of Poincaré’s
nomination for the Nobel Prize for Physics, when they promoted the recognition of
mathematical physics as a legitimate part of physics, and as we will see below for
the years following the sudden triumph of Einstein in late 1919.
8. THE UNINTENDED CONSEQUENCES OF EINSTEIN’S CELEBRITY
Following the conﬁrmation of Einstein’s general theory of relativity by the British
eclipse expedition of 1919, Einstein became an instant worldwide celebrity.54 I will
now argue that this event greatly contributed to the rewriting of the history of what
was now known as the “special” theory of relativity.55 As we have seen, no debate
ever erupted in the physics community before the War on the value of the relative
contributions of Poincaré and Einstein. Scientists did acknowledge Poincaré’s work
by citing his papers, but they saw a stronger connection between Lorentz and Einstein,
who both played a central and continuous role in the development of the electrodynamics of moving bodies. Now that Einstein was famous, and seemed to attract
all glory to himself, some scientists were irritated by a celebrity they felt distasteful
and found reasons to recall or ‘rediscover’ Poincaré’s role in a theory that was now
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
famous around the world.56
While the anti-Semitic feelings of some scientists lay dormant, the sudden
celebrity of Einstein could only awaken them and it is not surprising to ﬁnd that the
ﬁrst attempt to divorce relativity from its celebrated author came during a meeting in
Berlin in August 1920 of what Einstein called the “anti-relativity company”.57 It was
during this meeting that some speakers suggested that Einstein was a plagiarist. The
physicist Ernst Gehrcke had resurrected an old paper of 1898 by Paul Gerber on the
movement of the perihelion of Mercury that Einstein had not cited.58 A year later it
was the turn of another anti-relativist, Philip Lenard, to search for ‘precursors’, this
time connecting the famous formula E = mc2 to the name of Friedrich Hasenöhrl,
a non-Jewish Austrian physicist who died in the First World War and thus, as Max
Born noted, “could not object when his name was later misused to discredit Einstein’s
discovery”.59 For the prediction of the bending of light rays near the Sun, Lenard
found Johann Georg von Soldner who had made such a calculation on the basis of
Newtonian physics in 1801. Much later, in 1939, the physicist Wilhelm Lenz suggested to Max von Laue, who ﬁrmly rejected the proposition, that one could make
the theory of relativity more acceptable to the Third Reich by saying that it was ﬁrst
put forward by the Frenchman Henri Poincaré.60
Reasons other than strictly political or nationalistic can also play a role in the
search for ‘precursors’ of a celebrated theory. Disciplinary identity is one. In a
manner similar to that of Mittag-Lefﬂer who pushed Poincaré for the Nobel prize
in order to promote mathematical theory, this disciplinary bias was also probably at
work in 1921 when Oscar Klein urged Wolfgang Pauli to mention Poincaré’s papers
in his historical overview of relativity and to stress that another mathematician,
David Hilbert, independently found the basic equations of general relativity.61 Here
we have probably the ﬁrst clear struggle over the legitimate history of relativity that
involved Poincaré. It is interesting to see how the young Pauli, a physicist, skilfully
distinguished Einstein from Poincaré while dutifully following Klein’s advice. In the
ﬁrst section of Part 1 titled “Historical background (Poincaré, Lorentz, Einstein)”,
Pauli recalled that “there was ﬁrst of all Larmor who, as early as 1900, set up the
formulae now generally known as the Lorentz transformation” in his book Aether
and matter.62 The fact that Pauli mentioned Larmor provides an interesting contrast
to the debate around Poincaré. For it is indeed intriguing that contrary to Poincaré,
Larmor has not been the subject of any sustained movement championing his role
as ‘precursor’ of relativity, although in 1986 M. N. Macrossan spoke for a number
of physicists when he noted that “the credit for the ﬁrst presentation of the Lorentz
transformations, including the crucial time dilation, belongs to Larmor (1897)”, after
C. Kittel had observed in 1974 that “It has long appeared an historical anomaly that
Larmor’s work, which preceded Lorentz’s by four years, is so little known among
physicists”. But such statements have not given rise to anything like the passionate
debates surrounding Poincaré’s contributions to relativity.63
After having presented the Lorentz coordinate transformations that leave Maxwell
equations invariant, Pauli stressed “that even in this paper [of 1904] the relativity
principle was not at all apparent to Lorentz. Characteristically, and in contrast to
Einstein, he tried to understand the contraction in a causal way”.64 Then comes Pauli’s
characterization of Poincaré’s work: “The formal gaps left by Lorentz were ﬁlled
by Poincaré” and he “further corrected Lorentz’s formulae for the transformations
of charge density and current and also derived the complete covariance of the ﬁeld
equations of electron theory”. The next, and shorter, paragraph is devoted to Einstein:
“It was Einstein, ﬁnally, who in a way completed the basic formulation of this new
discipline. His paper of 1905 was submitted at almost the same time as Poincaré’s
article and had been written without previous knowledge of Lorentz’s paper of 1904.
It includes not only all the essential elements contained in the other two papers but
shows an entirely novel, and much more profound, understanding of the whole problem. This will now be demonstrated in detail.”65 It is quite clear that while recognizing
Poincaré’s work as Klein requested, Pauli stressed the fact that Einstein’s work is
“entirely novel” in its physical meaning. He presented a typical continuist narrative
which, while giving credit to many actors, led to Einstein’s novel synthesis, which
in turn erased the previous contributions.
This brings us to the famous visit of Einstein to Paris in April 1922, where the ghost
of Poincaré could not but hover over and appear in the discourses celebrating Einstein.
But even here it is striking that what the philosopher Xavier Leon, in his presentation
speech at the Société Française de Philosophie, stressed as Poincaré’s contributions
to the theory of electrons was the notion of electromagnetic momentum and what he
explicitly called the “Poincaré pressure”. As for Einstein himself, he “accomplished the
revolution that Poincaré had foreseen”.66 French newspapers also made a connection
between Einstein and Poincaré. The astronomer Charles Nordmann, a well-known
science popularizer who accompanied Einstein in Paris and covered his visit in detail,
remarked that the foundations of Einstein’s theory resided “for a large part on the ideas
of our Henri Poincaré”. As a promoter of Einstein, and given the delicate political
context of Franco-German relations, no doubt Nordmann wanted to make the theory
less German and more French.67 Despite these mentions of a connection between the
two great minds, no one presented their work as being equivalent. Even for the French
thinkers, Poincaré had only “foreseen” what Einstein “accomplished”.
Towards the end of his visit, Einstein received a letter from Gustave Le Bon,
the famous author of The psychology of the crowds and many other popular books,
complaining that the German physicist did not adequately acknowledge his own
priority concerning the equivalence of mass and energy. In July 1914, Le Bon had
ﬁrst published in the Comptes rendus of the Paris Academy of Science a statement
to the effect that the results Einstein mathematically obtained from the principle of
relativity were similar to those he arrived at between 1900 and 1907 in a series of
memoirs.68 His brief note having had no echo in the scientiﬁc community, Einstein’s
visit in Paris could only stimulate him to bring again his work to the attention of the
now celebrated physicist.
Probably unaware of the fact that Le Bon had a reputation for ﬁghting priority
disputes,69 Einstein exchanged several letters with him trying to explain (in French)
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
that he would be glad to refer to his works on this question if he could see where Le
Bon had given a proof of his assertion, as opposed to having discussed the simple
idea of equivalence. For whereas the general idea was very old, Einstein said, only
relativity theory had provided a proof of the equivalence including the identiﬁcation
of the proportionality factor as the square of the speed of light.70 In his exchanges of
letters, Einstein admitted that his “knowledge of the literature was relatively weak”71
and added — in response to Le Bon’s generalization concerning the “German habit
of totally ignoring foreign works” — that “crimes against intellectual property are
personal and not national affairs”. The exchanges seem to have ended after Le Bon
wrote on 19 July, citing the Bible, that “there is none more deaf than someone who
does not want to hear”. Though extreme, this brief affair nonetheless shows how
Einstein was attracting attention and creating envy around him. Moreover, as we
will see later (Section 11), the name of Le Bon resurfaced again in 1994 when a
Polytechnicien (as had been Poincaré) launched another attack on Einstein.
TRANSMUTING “LA DYNAMIQUE DE L’ÉLECTRON” INTO THE “THEORY OF
The ﬁrst clear statement that presents Poincaré’s work as a “génial précurseur” of
Einstein is provided by the Swiss physicist Edouard Guillaume, who, writing in
1924, ﬁnds it “incredible” that his memoir “is little known and almost never cited”.
Guillaume is thus the ﬁrst to mention the lack of citations of Poincaré’s paper.72 He
deplored the fact that the document was difﬁcult to ﬁnd, and commended GauthierVillars for making it available again through a new printing for which he wrote a
long introduction. But it is certainly more than a coincidence that this printing came
only after Einstein’s visit in Paris, in a context where relativity was everywhere and
many French books, some translated from German,73 presented the theory to a lay
audience without giving much credit to Poincaré. Even more interesting is the title
of the booklet: “The new mechanics: Conference, memoir and note on the theory
of relativity.”74 The new title thus suggested that what Poincaré in fact did was to
formulate independently another version of the “theory of relativity”, something
that was far from obvious to the non-expert through the original title of his papers
“On the dynamics of the electron”, a formulation that became obsolete shortly after
the rise of Einstein’s relativity theory. The book contains the 1909 conference “La
mécanique nouvelle” (pp. 1–17) ﬁrst presented at the Lille congress of the French
Association for the Advancement of Science and, later, in 1910, at the University of
Berlin.75 This is followed by the Rendiconti memoir “Sur la dynamique de l’électron”
(pp. 18–76), and ﬁnally by a reprint of the Note with the same title published in June
1905 in the Comptes rendus of the Paris Academy of Science (pp. 77–81). Guillaume
insists that Poincaré’s works on the new mechanics “are of the utmost importance”
as the “illustrious mathematician arrives at analytically identical results to those
adopted by the relativist school”.76
In promoting Poincaré, Guillaume was in fact doing two things at once: as a
foreigner he could call attention to the French contribution to relativity (hence the
title of the book) without being seen as “nationalist”, and he also used the occasion
to promote his own interpretation of the theory by enrolling Poincaré on his side.
For Guillaume believed in absolute time and this is why he promptly noted that
Poincaré’s route to relativity was not based “on the relativity of time and space” as
it was the case for the “relativist school”.77 He devoted about half of his introduction
to a presentation of his own peculiar interpretation of relativity, based on a varying
speed of light.
Guillaume had been battling against relativity for nearly ten years. An old friend
of Einstein, who like him was working at the Patent Ofﬁce, he even did experiments
in 1909 on the apparatus to measure small electric charges (the Maschinchen) that
Einstein had devised the year before.78 Guillaume had also translated into French
in 1910 some of Einstein papers for the Swiss Archives des sciences physiques et
naturelles.79 In 1917 he wrote to the French physicist Paul Langevin to present his
views and ask him to publish them in the Comptes rendus of the Paris Academy, but
Einstein corresponded with him in 1917 and then in 1920, trying to correct his
basic misunderstanding of relativity, but in vain.81 Guillaume’s observations were
“deﬁnitely a bit foul” Einstein frankly told him, adding to another common friend,
Marcel Grossman, that “Guillaume’s notice is stupid, like everything this man dashes
off about relativity”.82 Despite his friend’s negative comments, Guillaume continued
to promote his views and went to Paris during Einstein’s 1922 visit to confront him.
Newspapers reported in detail how he made a fool of himself in front of Einstein
Following Guillaume’s introduction to and republication of Poincaré’s papers, the
idea that Poincaré had not been adequately recognized for his contribution to relativity
was taken up the following year by the mathematician Paul Appell in his biography
of Poincaré. Referring to Guillaume, Appell insisted “on the fact that Poincaré has
been the precursor of the special theory of relativity”. Stressing that he did not want
to “open a discussion over the priority of this or that scientist in the ediﬁcation of
the new mechanics”, Appell concluded that it was sufﬁcient to have shown the part
played by his “compatriot in a theory that is now profoundly modifying our ideas
of the universe”.84
The theme of the ‘precursor’ would be taken up again by Louis de Broglie in his
talk on the occasion of the centenary of the birth of Poincaré in 1954. Speaking in
front of dignitaries at the Sorbonne, de Broglie, a Physics Nobel Prizewinner and
Secrétaire Perpétuel of the French Academy of Sciences, explained that Poincaré
“had all the elements of the theory of relativity” but was stopped from making the
last step provided by Einstein because of his “somewhat hypercritical turn of mind
or perhaps due to the fact that he was a pure mathematician”, which led him to a
nominalism that is ill-suited to physics. Like many before him, he mentioned the
“Poincaré pressure” as an important contribution to physics and concluded that “without Lorentz and Poincaré, Einstein would not have succeeded”.85 With his literary
gift, he summarized the relation between Poincaré and Einstein in a diplomatic way:
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
“the ﬁrst was the precursor of the theory of relativity while the second should be
considered as the true founding father of this theory.”86 For his part, General Dassault
from École Polytechnique insisted that Poincaré “had powerfully contributed to the
birth of relativity” and that “it is certain” that his Rendiconti memoir of 1906 “will
remain a classic in the history of the principle of relativity”.87
Other engineers88 did refer to the fact that Poincaré’s memoir predated Einstein’s,
but the consensus among the participants at the celebrations seems to have been
expressed by the Minister of National Education, who, echoing de Broglie, said that
“every one has his own peculiar genius” and that if Albert Einstein could build the
theory of relativity at the very moment that Poincaré came so near to discover it, it
was probably because in Einstein, the physicist dominated over the mathematician
whereas in Poincaré it was the other way around.89
THE IMPACT OF EDMUND WHITTAKER’S HISTORY OF THE THEORIES OF AETHER AND
While the participants at the Poincaré celebrations took an ecumenical approach
to the relative contribution of Einstein and their compatriot to the special theory of
relativity, despite the engineers’ attempts to push for more recognition, they seem
not to have been aware that across the channel, less than a year before, the English
mathematician Edmund Whittaker had advocated a more radical view promoting
Poincaré at the expense of Einstein in the second volume of his major book on
The history of the theories of aether and electricity, covering the period 1900–26.
Written in English and published in 1953, two years before the death of Einstein, it
appeared at a time when the new discipline of History and Philosophy of Science was
rapidly developing and in time for the ﬁftieth anniversary of the ‘birth’ of relativity.
These contingent conditions — the rise of a new discipline and the ﬁrst important
commemoration of relativity theory — seem to have converged to maximize the
impact of Whittaker’s thesis and make the topic of Poincaré’s role in the emergence
of relativity extremely timely.
Reviewers were quick to locate the “bomb” hidden in the 50-page-long narrative
of Chapter Two titled “The Relativity Theory of Poincaré and Lorentz”, in a 300-page
book mostly devoted to atomic physics, quantum theory and gravitation. Given that
Poincaré’s contributions were explicitly based on Lorentz’s earlier work, Whittaker’s
choice of the expression “Poincaré-Lorentz” instead of “Lorentz-Poincaré” (by analogy to the usual “Lorentz-Einstein”) probably reﬂects an unconscious disciplinary
bias in favour of the mathematician over the physicist. Going against a long tradition
making Einstein the major actor of what became the special theory of relativity, Whittaker asserted that Einstein’s role was in fact minor and that he simply reformulated
what Lorentz and Poincaré had already done. His famous 1905 paper only “set forth
the relativity theory of Poincaré and Lorentz with some ampliﬁcations”, though, he
added, it “attracted much attention”.90 Reviewing the book for Isis, the American
physicist Percy Bridgman noted that “the most controversial part of the book” will
be its “treatment of the special theory of relativity”, adding that “the cudgels in
favor of Einstein have already been taken up by several reviewers, including no less
a person than Max Born”.91
The book was fresh from the press when Max Born, a colleague of Whittaker
at the University of Edinburgh, wrote in alarm to his old friend Einstein that the
book “contains a history of the theory of relativity which is peculiar in that Lorentz
and Poincaré are credited with its discovery while your papers are treated as less
important”. Born added that he had done everything he could “during the last three
years to dissuade Whittaker from carrying his plan, which he cherished for a long
time and loved to talk about”. Born was very much annoyed because Whittaker was
“considered a great authority in the English speaking countries and many people are
going to believe him”.92 Although Einstein responded that he should not “lose any
sleep” over his friend’s book,93 Born reviewed it to make public his disagreements in
very diplomatic terms. Writing in the relatively new British journal for the philosophy of science — it had been founded in 1950 — Born ﬁrst insisted that the book
was a “magniﬁcent work” and that the only point of disagreement with its author
concerned the role of Einstein in the history of relativity. Noting that “the question
of attributing priority and connecting names to discoveries is a delicate one”, he
admitted that “much can be said in favour of Whittaker’s judgment”, recalling that
“from a mathematical standpoint the Lorentz transformations contain the whole of
special relativity”. Moreover, “there seems to be no doubt that Poincaré was, perhaps
a little ahead of Einstein, aware of most of the important physical consequences”.
That being said he asked: “But why has Einstein made such a great impression that
the others are forgotten by all but the specialists?” His answer was that Einstein,
starting from a deﬁnition of relative simultaneity using light signals, derived “the
Lorentz transformations and all their physical consequences”. He concluded that
“without disregarding the great contributions of Lorentz and Poincaré”, he sided
“with the general use in naming relativity after Einstein, if a name is desired at all”.94
In his obituary of Einstein published in the Biographical memoirs of Fellows of the
Royal Society in November 1955, Whittaker would repeat his view that Einstein
simply “adopted Poincaré’s Principle of relativity (using Poincaré’s name for it)”.95
He himself died a few months later, in March 1956. Even the author of his obituary
in the The British journal for the philosophy of science could not pass in silence
over this “affair”, and G. J. Whitrow wrote that Whittaker “did not do justice to the
originality of Einstein’s philosophy of time and his scientiﬁc method”, although
he added he could nonetheless understand “why as a historian he felt that it was
necessary to correct the widespread uncritical acceptance of the view that Einstein’s
contribution was unique”.96
An analysis of the many papers touching upon the ‘priority question’ shows that
the authors always cite Whittaker’s book and comment on it. Although rejecting his
extreme views, most agree that Poincaré played a role and they usually propose an
interpretation in terms of ‘precursors’ though not always using the term. In 1956
for example, in a paper titled “Fifty years of relativity” published in the American
magazine Science, P. G. Bergmann, who had participated the year before in the
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
international conference convened at Berne to celebrate the event, referred to Poincaré
and Whittaker in his paper but presented “the mathematician Poincaré” as having
formulated a principle of relativity for electromagnetic phenomena “without attempting a detailed physical analysis or interpretation. It remained for Einstein to provide
an integrated mathematical-physical analysis”.97 This narrative is rather typical of
the continuist conception of history spontaneously provided by physicists, which can
integrate many contributions without diminishing the weight given to the main actor:
Einstein. Max Born, who also participated in the Berne celebration, repeated that
he could not agree with Whittaker’s extreme position but added that it was “wrong
to forget these other men [Lorentz, Poincaré and Minkowski]”, because “relativity
was, after all, not a one-man discovery”.98 A few years later, in 1959, stimulated
by the reading of Whittaker’s “very interesting” book, the French physicist Theo
Kahan published in French a paper “On the origins of the special theory of relativity” deploring the “vain or sordid questions of priority” and reafﬁrming Einstein’s
powerful “élan spirituel” which gave rise in the space of ten years to special and
general relativity.99 The philosopher Adolph Grünbaum considered for his part that
Whittaker’s “disparaging evaluation of Einstein’s role” was simply an illustration
of his philosophical misconceptions and concluded with the wish that the history of
relativity be “tackled by historians having the intellectual command of the philosophical foundation of that theory”.100
In the mid-1960s, Whittaker found a successor in the philosopher G. H. Keswani,
who published a three-part paper on “The origin and concept of relativity”, in which
he reformulated in more historical details Whittaker’s thesis that “Poincaré postulated
the principle of relativity, which Einstein learnt and adopted using Lorentz’s mathematical formulation”.101 Reacting promptly to Keswani’s analysis, the philosopher
Herbert Dingle noted that, as a mathematician, Whittaker was “particularly susceptible to the danger of judging a theory by its mathematical form alone”, and insisted
that the difference between theories lies “not in the mathematics but in the physical
ideas which the mathematics represent”.102 A few months later, Karl Popper also
intervened in the debate to conclude that although “Poincaré has to be credited with
predicting the coming of what is rightly called Einstein’s theory of relativity, it was
Einstein who achieved it”.103
In parallel to the debate between philosophers, scientists were also attracted by
Whittaker’s controversial thesis and wrote a series of papers addressing the question of Poincaré’s contribution to relativity in the American journal of physics, a
pedagogical journal giving much place to history of science as a tool for physics
teachers. In 1964, Charles Scribner invoked pedagogical arguments for a continuist
history. In a paper titled “Henri Poincaré and the principle of relativity”, he noted
that in elementary presentations of the special theory of relativity, Einstein was too
often seen as the unique source of the theory, a situation “apt to create pedagogical
difﬁculties, for it is much easier to understand Einstein’s accomplishment if one sees
how he was able to create an original synthesis containing a number of ideas which
had already been introduced by other mathematical physicists in their own efforts
to create a satisfactory theory of electrodynamics”.104 Opposing Whittaker’s view as
well as the one centred on Einstein as failing to “do justice to the available historical
evidence”, he analysed in detail Poincaré’s contributions between 1895 and 1905.
In 1965, the physicist H. M. Schwartz wrote that the publication of a translation
of the 1906 Rendiconti paper would serve “to do justice to Poincaré’s part in the
development of the theory of relativity”, the part played by Lorentz, Einstein and
Minkowski being already documented “in the well-known collection of their original
publications”.105 Schwartz did deliver his partial translation of selected sections six
years later, just after another physicist, C. W. Kilmister, a professor of mathematics at King’s College, London, had also provided a partial translation in a volume
dedicated to “classic” texts on the special theory of relativity.106 So, over the period
1965–75, there was an intense debate about the extent to which Poincaré did or did
not ‘anticipate’ or even ‘discover’ the theory attributed to Einstein.107 Finally, in 1983
Kilmister joined forces with Keswani to publish an English translation of Poincaré’s
June 1905 Comptes rendus paper.108
From the mid-1970s, philosophers interested in relativity seem to have turned their
back on that debate and essentially limited their discussions to ‘rational reconstructions’ of Einstein’s path to relativity, based on close conceptual and philosophical
readings of original papers.109 By that time, Poincaré’s contributions to relativity
had been thoroughly discussed by historians and philosophers of science as well
as by historically-minded scientists. As a consequence, the entry on Poincaré in the
authoritative Dictionary of scientiﬁc biography could signal that “many physicists
consider that Poincaré shares with Lorentz and Einstein the credit for the invention
of the special theory of relativity”.110
Ironically, even if the vast majority never accepted Whittaker’s view, one can point
to his bold and radical “marginalization” of Einstein as the major cause of the resurrection of “Poincaré the physicist” as an actor worthy of analysis by historians and
philosophers of physics. For even if the dominant narrative stayed centred on Einstein
as the “revolutionary” physicist, most now admit that important contributions were
made by Poincaré, in addition to those of Lorentz, who was already acknowledged
by the traditional narrative of relativity.
It is intriguing to observe that when they lack a disciplinary bias of the sort
exhibited by mathematicians such as Klein and Whittaker, many of the promoters
of Poincaré were either sceptical of relativity or opposed to it. Hence Keswani, who
wrote papers endorsing Whittaker’s view, was convinced that Einstein’s special
relativity was inconsistent,111 and we have seen that Guillaume was opposed to the
relativity of time. The American physicist Herbert E. Ives, also critical of Einstein’s
version of relativity, even talked about “the Lorentz-Poincaré transformations”,112
and Whittaker himself was a promoter of Eddington’s version of general relativity.113
This is not, of course, to suggest that it is not legitimate to agree (or disagree) with
their point of view but to insist that the actors at play in the game of ﬁnding precursors were not randomly distributed but had many common characteristics and were
actively rewriting history in conformity with their interests. From this point of view,
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
Whittaker’s book, despite its erudition, is less a distanced contribution to the history
of science than an essay by a mathematician aiming at redistributing symbolic credit
among scientists in favour of mathematicians. And in a symmetric manner, most of
the responses to his book were framed as a “defence” of Einstein’s priority.
11. TRANSFORMING THE PUBLIC IMAGE OF POINCARÉ
All the discussions surveyed above were essentially going on between experts in the
scientiﬁc ﬁelds of physics and in history and philosophy of science, and had little
impact among the general population, among whom “relativity” was (and still is)
synonymous with “Einstein”, who became a brand name in itself independent of his
ﬁrst name, Albert. Now that all the scholarly elements were at hand, it was possible
to take a further step and try to change the public image of Poincaré by proposing a
more radical interpretation of the reasons behind Einstein’s domination. Three commemorations would provide the necessary context for those who still thought that
Poincaré had somehow been sidelined and never received proper recognition from
‘posterity’: the bicentenary of the École Polytechnique of Paris in 1994, the 150th
anniversary of Poincaré’s birthday in 2004, and ﬁnally, the UNESCO World Year
of Physics in 2005. But instead of the ecumenical and civilized approach that had
dominated the discussions until then, a more aggressive tone would prevail, some
even suggesting plagiarism on the part of Einstein. With such ‘new’ and sulphurous
elements, the media would be more willing to cover what had been, until then, a
mainly academic debate.114
In April 1994, on the occasion of the bicentenary of the Polytechnique, the journal
of the fraternity of Polytechnicians, La jaune et la rouge, published a paper by the
engineer Jules Leveugle, a 1943 graduate and member of that very cohesive group of
Anciens élèves de l’X (as the Polytechnique is usually called by its members), simply
titled “Poincaré and relativity”. Conscious of the delicate nature of the content of
the paper, the editor of the journal wrote a foreword explaining that it was only after
much hesitation and “after having obtained satisfactory responses to all their objections” that they decided to print Leveugle’s paper, which “many eminent physicists
and former X had encouraged [him] to publish”, in the section “Libres propos”.115
Leveugle argues that the theory of relativity has been completely developed,
including the famous equation E = mc2, in Poincaré’s publications, all of them prior
to Einstein’s 1905 papers. For professional historians of physics, there was nothing
new in this paper, except the peculiar interpretation, the strongly nationalist tone
and the insinuations of plagiarism. Denouncing the fact that French historians of
science “have ignored the originality and anteriority of Poincaré in the genesis of
relativity theory”, Leveugle insisted that it was high time, given the 200th anniversary of the École Polytechnique, that a movement took form to correct the injustice
done to Poincaré, their most illustrious colleague who discovered relativity before
Einstein. One had to “do something so that teachers, textbooks and media note the
originality and anteriority of Poincaré’s ideas, of his contribution to the elaboration
of the famous formula E = mc2, of the scientiﬁc, pedagogic and moral value of the
enduring dialogue between Poincaré and Lorentz, which led to the discovery of
relativity theory”.116 He also proposed to name the principle of relativity “Poincaré
Principle”, and “Poincaré-Planck” the “mythic” equation relating inertia to energy.
Finally, adding a French name to the usual ‘precursor’ of that equation (Friedrich
Hasenöhrl), he recalled the contributions of Gustave Le Bon (discussed in Section
8) and the criticisms of Einstein’s demonstrations of the equivalence between inertia
and energy by the American physicist H. E. Ives.117
Circulating essentially among Polytechnicians, Leveugle’s paper seems not to
have had any public impact until ten years later, when its author re-published it as
a book-length analysis in the more favourable context of the 150th anniversary of
Poincaré’s birth (2004) and the coming UNESCO 2005 World Year of Physics. Leveugle’s book118 ﬁnally hit the newspapers with its provocative thesis. Though most were
sceptical and did not want to give high visibility to a book that showed restrained but
nonetheless visible hatred towards Einstein, the promotion of Poincaré was ﬁnally
in the media. With headlines like “Einstein plagiarist?” and “Einstein relativized”,
the major newspapers and magazines covered the new debate.119
The context provided by the 2005 Year of Physics gave an unprecedented public
visibility to Einstein as the uncontested incarnation of the discipline in the public
imagination. In France however it offered a chance to promote “Poincaré the physicist”, mainly, as we have just seen, through the polemic of the ‘paternity’ of relativity. Given the logic of media, where only ‘scandalous’ views and truly newsworthy
events can get off the ground, it can be said that it was only with extreme viewpoints
that one could make Poincaré’s role in physics known outside the circle of specialists. Such an intense commemoration could only stimulate those who thought that
Poincaré was somehow unjustly forgotten, if not in the specialized literature at least
in the public imagination.
In addition to popular books, which can reach a large public, a powerful tool
to construct collective memory, particularly among high school science students,
is provided by documentary ﬁlms. It is thus signiﬁcant that in 2005 a 26-minute
documentary on Poincaré was produced with high school students as the intended
audience. Titled “Tout est relatif Professeur Poincaré!” (translated as “It’s all relative Professor Poincaré!”), it exists not only in French but in the English, German
and Chinese languages, the latter versions probably serving the promotion of French
culture abroad and of Poincaré as an important incarnation of it.120 Its basic message
is that Poincaré’s paper on the dynamics of the electron resembles in many respects
Einstein’s 1905 paper on relativity but that this ‘fact’ is ignored by the general public.
The ﬁlm promotes a counter-narrative and counter-memory to the dominant one centred on Einstein, in order to rewrite history and replace what the narrator curiously
calls “popular history” by another one in which Poincaré will ﬁnally have a place
“in the Pantheon of Scientiﬁc Greats”.121
As underlined by Christopher Flood, “the political dimension of cultural memory
varies in function of the ideological environments constituted by different subcultures”, and “representation of the past is an object of competition between different
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
subcultures which naturally wish to shape it particularly to their own normative
vision”.122 Though the author focused on the memory of political events, it is clear
that the analysis is also valid for scientiﬁc events, even though they are less likely
to provoke debates as heated as those usually accompanying revisionist histories of
major social events like revolutions and wars.
12. THE SOCIAL FUNCTION OF ‘PRECURSORS’
As many scientists, philosophers and historians have used the term ‘precursor’ to
qualify Poincaré’s role in the genesis of relativity theory, it is worth looking more
closely at the uses of this notion by historians of science and scientists. In the discipline of history of science, George Canguilhem brilliantly argued that the notion
of ‘precursor’ is an artifact, “a counterfeit historical object”.123 One of the practical
effects of the autonomy of the history of science from the practice of science is the
elimination of what J. T. Clark called “the virus of the precursor”.124 Another founding father of the discipline, Alexandre Koyré, had also noted with irony that nobody
has ever regarded himself as a ‘precursor’ of someone else, nor been able to do so,
since the notion makes no historical sense except, to use the words of Canguilhem,
by “substituting the logical time of truth relations for the historical time of these
relations’ invention”, thus treating “the history of science as though it were a copy
of science and its object a copy of the object of science”.125
While these philosophical and methodological arguments should be sufﬁcient to
eliminate the notion of (and the search for) precursor from the discipline of history of
science, thus effectively blocking most useless debates about ‘priority’, it is important
to note that the notion of precursor does make sense for a scientiﬁc community that
must construct its own ‘ofﬁcial history’ accommodating the work of all its members. As we have already seen several times, scientists tend to construct continuist
histories, and precursors serve a precise social function in the construction of the
collective memory of a community, by creating links that provide continuity in time
and between successive generations of scientists. This spontaneous philosophy of
history as uniform progress is fundamentally ecumenical and thus minimizes potential
tensions between the different groups and specialties forming the discipline. It is thus
not surprising to see that the term often appears in actors’ and scientists’ histories of
science. Instead of confronting scientists with each other, it is simpler for the harmonious relations within and between disciplines to construct a global scenario uniting
the contributions of everyone in a narrative of the continuous progress of science.
Once the social function of precursors in historical narrative written by actors of
a discipline is acknowledged and distinguished from the approach of the discipline
of the history of science, it becomes easier to understand that the notion is frequent
among scientist-historians and much less used by philosopher-historians and even
less by sociologist-historians. Canguilhem’s analysis is also important in that it makes
visible the fact that most debates about the speciﬁc contributions of Einstein and
Poincaré are an effect of the confusion of roles where historians forget their speciﬁc
method and object and become actors in the game of attributing ‘proper credit’,
instead of analysing it.126 In other words, the “all-too-vexed question” (to use Peter
Galison’s expression127) of Einstein’s relation to Poincaré is better disposed of not
by generously but arbitrarily attributing ‘equal credit’ to the actors, but by historicizing thoroughly the very notion of credit attribution, while leaving its practice to the
members of the tribe in the scientiﬁc ﬁeld.
In his analysis of the discussions surrounding the role of Poincaré in the creation of
the special theory of relativity, Darrigol distinguished the approach of the “physicisthistorian” from that of the “philosopher-historian”.128 There is a third possibility:
that of the “sociologist-historian” adopted here. I think this approach is best suited
to solve the “mystery of the Einstein-Poincaré connection”. Instead of concentrating
my attention on the internal logic of the texts of the authors, I have looked at the
manner they were read and cited (or not) by others, and I have found that there was
a general tendency on the part of the physics community to talk of relativity in terms
of Lorentz’s and Einstein’s contributions. They did not ‘neglect’ Poincaré but saw in
his works a complement to those of Lorentz.
Darrigol concluded his analysis by saying that “for the unprejudiced historian, it
should be clear that in the German culture of the ﬁrst years of relativity, the PoincaréLorentz and the Einstein views had equal chances”.129 Yet from the unprejudiced
sociologist-historian’s point of view developed here, it seems clear that there simply
was no such real choice for the actors of the time: (1) physicists never talked about a
“Poincaré-Lorentz” theory, whereas they frequently talked about “Lorentz-Einstein”,
and (2) they rarely cited Poincaré’s two papers in their publications over the period
1905–39. In other words, from an historicist point of view — which is probably the
only practical way of being ‘unprejudiced’ — the actors of the time have never perceived and never discussed a ‘choice’ between Einstein and Poincaré, though they have
much discussed the link between Lorentz and Einstein. To paraphrase Canguilhem,
such a choice exists only in the space of logical possibilities, for as I have shown, all
the discussions concerning Poincaré were a reaction to Einstein’s having become a
famous physicist after 1919, thus stimulating the writing and rewriting of the history
of that ‘revolution’.
It should thus come as no surprise that, having implicitly accepted the formulation of a question framed by scientists in terms of evaluating who should get ‘proper
credit’, Darrigol simply rediscovered the ‘solution’ already proposed to that “vexed
question” by Max Born ﬁfty years earlier. As we have seen, this question of priority
was widely discussed among scientists and philosophers of science in the decade
1955–65 and reappeared in 1994 with the celebration of the 200th anniversary of
Polytechnique and ﬁnally culminated in 2005 with the UNESCO World Year of
Physics when the debate hit the French newspapers.
As is often the case in history, nobody wants to be associated with a ‘loser’ but
once it has become obvious that there is a clear ‘winner’ then it is often observed
that some people have interests to retrospectively ‘discover’ that someone else had
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
also made an equivalent contribution that went unnoticed at the time.130 That is a
basic sociological reality, as unpleasant as it may seem from a moral point of view.
For whatever reasons one may have to try to redistribute ‘credit’ among actors in
the ﬁeld of “relativity”, be they political, nationalist, philosophical, conceptual, or
otherwise: they have all to do with our present historical and memorial predicaments
and nothing to do with the formative period of the theory between 1905 and 1911.
As analysts, historians should refrain from taking side in this artiﬁcial “debate” and
keep their “ethnographic eyes” wide open. Their attitude should not be to try to
decide if such and such a person was or not really a ‘precursor’ — thus losing their
autonomy and becoming scientists’ advocates131 — but to take the very ‘question of
priority’ as being at stake in the scientiﬁc ﬁeld itself in the recurrent struggle over
the legitimate history of a ‘discovery’.
I would like to thank Jean Eisenstaedt, Jeffrey Crelinstein, Camille Limoges, Aant
Elzinga, the editor of this journal and his reviewers for their useful comments and
suggestions. I also thank Jutta Schickore for help with the German texts and my
research assistants Vincent Larivière and Matthew Wallace for the preparation of
the bibliometric data.
1. For a series of papers on commemoration in science see Pnina G. Abir Am and Clark A. Elliott
(eds), Commemorative practices in science: Historical perspectives on the politics of collective
memory (Osiris, xiv (1999)); and Pnina G. Abir Am (ed.), La mise en mémoire de la science:
Pour une ethnographie historique des rites commémoratifs (Paris, 1998).
2. The debate on the role of Poincaré in the emergence of relativity theory has been particularly hot in
France where amateurs have published books claiming that Einstein plagiarized Poincaré. For
a defence of Einstein against this ‘revisionist’ literature, see Roger Cerf, “Dismissing renewed
attempts to deny Einstein the discovery of special relativity”, American journal of physics, lxxiv
(2006), 818–24. For an analysis of the events, see below Section 11, and Yves Gingras, “Henri
Poincaré: The movie. The unintended consequences of scientiﬁc commemorations”, Isis, xcviii
3. O. Darrigol, “The mystery of the Einstein–Poincaré connection”, Isis, xcv (2004), 614–26, p.
619, emphasis in the text. As we will see in Section 10, the view that Einstein is not the only
contributor to relativity was already promoted by Max Born in the mid-1950s in the context
of Edmund Whittaker’s famous thesis making Poincaré and Lorentz the sole ‘discoverers’ of
relativity theory. In making the same suggestion in an earlier paper, Darrigol mentioned, as an
example to follow, the case of quantum mechanics where it is usual to talk of Heisenberg and
Schrödinger as both creators of the theory. From the historicist point of view adopted here,
however, the example is not well suited as co-citation measure clearly shows that physicists at
the time (that is, between 1926 and 1930, for example) immediately connected the two authors,
whereas, as we will see, this was rarely the case for Poincaré and Einstein between 1905 and
1912. The two cases are thus not analogous from an historical point of view; see O. Darrigol,
“The electrodynamic origin of relativity”, Historical studies in the physical and biological
sciences, xxvi (1996), 241–312, p. 312.
4. R. K. Merton, The sociology of science: Theoretical and empirical investigations (Chicago, 1973),
5. Rob Iliffe, “‘In the warehouse’: Privacy, property and priority in the early Royal Society”, History
of science, xxx (1992), 29–68, p. 30.
6. It should be noted that the name of Lorentz has never been at stake in the debate over the creators of
relativity, the theory being often called, as early as 1906, “Lorentz-Einstein”. See Gerald Holton,
Thematic origins of scientiﬁc thought: Kepler to Einstein (Cambridge, MA, 1973), 234–5; and
Christa Jungnickel and Russell McCormmach, Intellectual mastery of nature, ii: The now mighty
theoretical physics, 1875–1925 (Chicago, 1986), 248–51.
7. Pierre Bourdieu, “Sur les rapports entre la sociologie et l’histoire en Allemagne et en France”, Actes
de la recherche en sciences sociales, cvi–cvii (1995), 108–22, p. 117.
8. For a review and extensive bibliography on citations see Loet Leydesdorff, “Theories of citation?”,
Scientometrics, xliii (1998), 5–25. On co-citation analysis see H. Small and B. C. Grifﬁth,
“Structure of scientiﬁc literatures. 1. Identifying and graphing specialties”, Science studies, iv
(1974), 17–40; H. G. Small, “Co-citation model of a scientiﬁc specialty — Longitudinal study of
collagen research”, Social studies of science, vii (1977), 139–66; and H. Small, “Cited documents
as concept symbols”, Social studies of science, viii (1978), 327–40.
9. For examples of uses of this indicator for relativity, see Jozsef Illy, “Revolutions in a revolution”,
Studies in history and philosophy of science, xii (1981), 173–210; Hubert F. M. Goenner, “The
reception of the theory of relativity in Germany as reﬂected by books published between 1908 and
1945”, in J. Eisenstaedt and A. J. Kox (eds), Studies in the history of general relativity (Boston,
1993), 15–38; and Walter Scott, “Minkowski, mathematicians, and the mathematical theory of
relativity”, in H. Goenner, J. Renn, J. Ritter and T. Sauer (eds), The expanding worlds of general
relativity (Boston, 1999), 45–86.
10. Darrigol, “The electrodynamic origin of relativity” (ref. 3), 311.
11. Andrew Warwick, Masters of theory: Cambridge and the rise of mathematical physics (Chicago,
12. On this task see Darrigol, “The mystery” (ref. 3), 619. A truly historicist approach also has
methodological consequences: while internal analysis tends to consider all texts on the same
footing as if they were equivalent and accessible to all potential readers, the historicist approach
recognizes that a textbook is a different object from a paper in a scientiﬁc journal, and that an
unpublished course has not the same visibility in the scientiﬁc ﬁeld as a paper in a well-known
journal. Though obvious to historians, these details are often forgotten in many conceptual
analyses of Poincaré’s work in physics. In other words, to be an historical object, a document
must be known to at least some actors of the time.
13. For examples of such brief statements, see R. Staley, “On the histories of relativity: The propagation
and elaboration of relativity theory in participant histories in Germany, 1905–1911”, Isis,
lxxxix (1988), 263–99, p. 277; M. Paty, Einstein philosophe (Paris, 1993), 43; and Michel
Janssen, “Reconsidering a scientiﬁc revolution: The case of Einstein versus Lorentz”, Physics
in perspective, iv (2002), 421–46, p. 428. For a more detailed analysis of why Minkowski did
not cite Poincaré’s 1906 paper see Scott, op. cit. (ref. 9), 56–59.
14. If one starts with the idea that Poincaré’s project was different from that of Einstein, there is no
reason to expect that Poincaré should cite Einstein and thus no “surprise” or “scandal” that he
does not. That is the view adopted by the physicist-historian Abraham Pais. See A. Pais, ‘Subtle
is the Lord...’: The science and life of Albert Einstein (Oxford, 1982), 167.
15. On such an approach see Jules Leveugle, La relativité, Poincaré et Einstein, Planck, Hilbert: Histoire
véridique de la théorie de la relativité (Paris, 2004), and Section 11 below.
16. For details on the Century of science database and the list of journals see http://scientiﬁc.thomson.
com. Some journals have merged over time or changed titles.
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
17. Given that citations to our authors (Poincaré, Einstein and Lorentz) are concentrated in physics
and mathematics, with some chemistry (for Brownian motion and speciﬁc heat of solids for
example), we can limit our analysis to these journals, though our search included all journals
in the database over the period 1900–79. The other ﬁelds covered in the period 1900–44 are
medicine (100 journals), biology (43), engineering (8), geology (5), and psychology (20). We
have assigned discipline on the basis of the title of the journal. For the period 1945–79 the data
come from Thomson Scientiﬁc Web of Knowledge and include science as well as social sciences
and humanities journals.
18. E. T. Whittaker, History of the theories of aether and electricity, ii: The modern theories, 1900–1926
19. Alhough not all scientiﬁc journals are covered, the major physics journals are included and it is
highly improbable that adding other journals would signiﬁcantly change the trends. Hence, for
the period 1905–11, data for the British journal Philosophical magazine are missing, but we have
checked the papers related to “relativity” in this journal and they too tend not to cite Poincaré.
We have also checked comparable articles in the Italian journal Il nuovo cimento with the same
results. These papers have been identiﬁed using the online database of the Archivio di Storia
della Fisica. See http://ﬁsicavolta.unipv.it/asf/archives.asp.
20. Pierre Bourdieu, Science of science and reﬂexivity (Chicago, 2004).
21. Let us recall that two documents (or two authors) are said to be “co-cited” when they both appear
in the references of another paper. Thus, the number of co-citations to two authors (or papers)
in a particular year is given by the number of different papers which, in that year, cite them
22. As early as 1972, N. L. Balàsz had suggested looking at actual citations to discuss the case of
Poincaré and Einstein. Using Science abstracts between 1905 and 1910, he concluded, on
the basis of fewer than a dozen papers, that “the contemporary scientiﬁc community did not
attribute the construction of the theory of relativity to Poincaré”. As we will see, we agree with
this conclusion on the basis of the analysis of thousands of citations and co-citations, using a
tool that did not exist when Balàsz wrote his paper. N. L. Balàsz, “The acceptability of physical
theories: Poincaré versus Einstein”, in L. O’Raifeartaigh (ed.), General relativity: Papers in
honour of J. L. Synge (Oxford, 1972), 21–34.
23. For an analysis of the limits of citation analysis see David Edge, “Quantitative measures of
communication in science: A critical overview”, History of science, xvii (1979), 102–34.
24. Richard H. Beyler, “The physics community in the national socialist era”, in Renn J. (ed.), Albert
Einstein: Chief engineer of the universe (Berlin, 2005), 320–3, p. 322; and A. I. Miller, “A précis
of Edmund Whittaker’s ‘Relativity theory of Poincaré and Lorentz’”, Archives internationales
d’histoire des sciences, xxxvii (1987), 93–103, p. 95.
25. Merton, op. cit. (ref. 4).
26. Einstein cited Poincaré’s 1900 paper published in the Lorentz Festschrift, in his second paper on the
inertia of energy, published in Annalen der Physik in 1906 in which he also cites his own ﬁrst
paper on inertia of energy, thus creating a co-citation “Einstein-Poincaré”. See The collected
papers of Albert Einstein (Princeton, 1987- ; hereafter CPAE), ii, English translation, Document
35, 200. As is well known, Poincaré never cited Einstein.
27. Of course, scientists could also send reprints as Poincaré did to Lorentz, but it was certainly more
efﬁcient to publish in well-known journals to reach a larger audience, as Einstein did by sending
all his early papers to Annalen der Physik, the best physics journal of the time. On the evolution
of the content of this journal at the turn of the century, see Jungnickel and McCormmach, op.
cit. (ref. 6), 309–23.
28. We can limit our inquiry to the period 1905–11, usually accepted as the formative period of
relativity. As is well known, Arnold Sommerfeld in 1911 considered that relativity was a secure
part of physics, while in 1913 Lorentz observed that the acceptance of Einstein’s concepts was
surprisingly rapid. Also, given that Poincaré died in 1912 and that from that time on Einstein was
more concerned with the development of general relativity, this time frame makes it possible to
look at the rise of Einstein before he became famous and to compare his citation pattern to those
of well recognized ﬁgures in their ﬁelds, Poincaré and Lorentz. The data include self-citations.
On network analysis and degree of centrality, which measures the number of links connecting an
actor to all the others, see L. C. Freeman, “Centrality in social networks: Conceptual clariﬁcation”,
Social networks, i (1978/79), 215–39; and S. Wasserman and K. Faust, Social networks analysis:
Methods and applications (Cambridge, 1994). Note that whereas social network analysis usually
concentrates on social links between people, here we apply the technique to co-citations links,
which have a more conceptual nature. For a recent survey of co-citation analysis, see Markus
Gmür, “Co-citation analysis and the search for invisible colleges: A methodological evaluation”,
Scientometrics, lvii (2003), 27–57.
For a more detailed analysis see Yves Gingras, “Mapping the changing centrality of physicists
(1900–1944)”, Proceedings of ISSI 2007 11th International Conference of the International
Society for Scientometrics and Informetrics, Madrid, Spain, 2007, 314–20.
On Lorentz see Tetu Hirosige, “Origins of Lorentz’s theory of electron and the concept of the
electromagnetic ﬁeld”, Historical studies in the physical sciences, i (1969), 151–209; and Nancy
J. Nersessian, “Why wasn’t Lorentz Einstein? An examination of the scientiﬁc method of H. A.
Lorentz”, Centaurus, xxix (1986), 205–42.
The letter, in French, is reproduced in Arthur I. Miller, Albert Einstein’s special theory of
relativity: Emergence and early interpretation (1905–1911) (Reading, MA, 1981), 336–37;
H. Poincaré, “Sur la dynamique de l’électron”, in H. Poincaré, La mécanique nouvelle: Conférence,
mémoire et note sur la théorie de la relativité, with introduction by Édouard Guillaume (Paris,
1924), 18–76, p. 19; we use the translation by H. M. Schwartz, “Poincaré’s Rendiconti paper
on relativity. Part I”, American journal of physics, xxxix (1971), 1287–94, p. 1288, emphasis
See E. Crawford, The beginnings of the Nobel Institution: The science prizes, 1901–1915 (Cambridge,
A little exercise in reﬂexivity should easily convince the reader that this mechanism is also at work
in the discipline of history of science….
At the request of the Swedish mathematician Gösta Mittag-Lefﬂer, Poincaré wrote in 1901 a long
report on his own works. He presented his own contributions to electrodynamics in a section
titled “Critique des théories physiques”. See Oeuvres de Henri Poincaré (Paris, 1954), ix, 7–14,
H. A. Lorentz, The theory of electrons, 2nd edn (New York, 1952; lst pub. 1915), 213–15.
Ibid., Preface, dated January 1909.
H. A. Lorentz, Problems of modern physics, ed. by H. Bateman (Boston, 1927), 127 and 133. Others
noticed Poincaré’s search for invariants. See for example, E. Cunningham, The principle of
relativity (Cambridge, 1914), 173.
H. A. Lorentz, untitled address, The astrophysical journal, lviii (1928), 345–51, p. 350.
The editors of the letter note that Einstein’s name is written on the draft letter but does not appear
in the typed copy. See http://www.univ-nancy2.fr/poincare/chp/.
Mittag-Lefﬂer to Poincaré, 5 July 1909, cited in Scott, op. cit. (ref. 9), 57.
Darboux et al. to the Nobel Committee, c. 1 January 1910, in La correspondance d’Henri Poincaré,
electronic version at http://www.univ-nancy2.fr/poincare/chp/.
V. Volterra to Nobel Committee, 10 January 1910, ibid. Incidentally, the same Volterra had given
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
a series of lectures at Clark University in 1909 in which he talked about relativity theory,
commenting on Minkowski’s “profound memoir” which “showed in a new light the ideas of
Lorentz and those of Einstein on the relations between space and time”. He also mentioned the
“relativity theorem of Lorentz”. At the end of his lecture he refered to Poincaré’s Rendiconti paper,
calling attention only to the fact that he showed that the action is invariant under the Lorentz
transformations. See Vito Volterra, Ernest Rutherford, Robert Williams Wood and Carl Carus,
Lectures at the celebration of the twentieth anniversary of the foundation of Clark University
(Worcester, MA, 1912), 13, 22, 27.
For details on this story see Crawford, op. cit. (ref. 34), 136–49.
Pais, op. cit. (ref. 14), 505; Elizabeth Crawford, J. L. Heilbron and Rebecca Ulrrich, The Nobel
population, 1901–1937 (Berkeley, 1987), 52–53; and Aant Elzinga, Einstein’s Nobel prize: A
glimpse behind closed doors. The archival evidence (Sagamore Beach, MA, 2006). Lorentz had
already received the Nobel Prize for Physics in 1902 with Pieter Zeeman.
On Langevin, see Camillo Cuvaj, “Paul Langevin and the theory of relativity”, Japanese studies in the
history of science, x (1971), 113–42; M. Paty, “Poincaré, Langevin et Einstein”, Épistémologiques,
ii (2002), 33–73; and “Paul Langevin (1871–1946), la relativité et les quanta”, Bulletin de La
Société Française de Physique, cxix (1999), 15–20.
P. Langevin, “L’œuvre d’Henri Poincaré: Le physicien”, Revue de métaphysique et de morale,
Supplément, xxi/5 (1913), 675–718, pp. 699 and 702.
Gösta Mittag-Lefﬂer to Einstein, 16 December 1919, CPAE, ix, English translation, Document
The invitation arrived at a time when Einstein was totally taken up by travel and the obligations
imposed by his new celebrity, and he replied to Mittag-Lefﬂer that he could not ﬁnd time to
“write the planned essay on Poincaré’s position on the problem of geometry and experience”.
See Einstein to Mittag-Lefﬂer, 21 July 1920, CPAE, x, English translation, Document 79, 212.
Note that his planned essay looked more like a discussion of the philosophy of conventionalism
than a physical discussion of “the relationship between space, matter, and time” as suggested
W. Wien, “Die Bedeutung Henri Poincaré’s Für Die Physik”, Acta mathematica, xxxviii (1921),
289–91, p. 290.
H. A. Lorentz, “Deux mémoires de Henri Poincaré sur la physique mathématique”, Acta mathematica,
xxxviii (1921), 293–308; reprinted in Oeuvres de Henri Poincaré (ref. 36), ix, 683–95.
M. Paty, “The scientiﬁc reception of relativity in France”, in T. F. Glick (ed.), The comparative
reception of relativity (Dordrecht, 1987), 113–67, p. 119.
For an analysis of this event see Alistair Sponsel, “Constructing a ‘revolution in science’: The
campaign to promote a favourable reception for the 1919 solar eclipse experiments”, The British
journal for the history of science, xxxv (2002), 439–67. See also Marshall Missner, “Why Einstein
became famous in America”, Social studies of science, xv (1985), 267–91.
For the early histories of relativity up to 1911 see Richard Staley, “On the histories of relativity: The
propagation and elaboration of relativity theory in participant histories in Germany, 1905–1911”,
Isis, lxxxix (1998), 263–99.
On the irritation of some of his friends at the “cult of personality” developing around Einstein see Max
Born, The Born–Einstein letters, 1916–1955, new edn (New York, 2005), 32–43; and Lewis Elton,
“Einstein, general relativity and the German press 1919–1920”, Isis, lxxvii (1986), 95–103.
CPAE, vii, English translation, Document 45, 197.
On this debate, see Hubert Goenner, “The reaction to relativity theory, I : The anti-Einstein campaign
in Germany in 1920”, Science in context, vi (1993), 107–33 ; M. Crowe, “Einstein’s encounters
with German anti-relativists”, in CPAE, vii, 101–13; and Milena Wazeck, “Einstein in the daily
press: A glimpse into the Gehrcke papers” (Max Planck Institute for the History of Science,
Preprint 271, 2004), 67–85.
Max Born, Physics in my generation (New York, 1969), 105–6. For a history of that equation, see W.
L. Fadner, “Did Einstein really discover ‘E = mc2?’”, Journal of physics, lvi (1988), 114–22.
For details see A. D. Beyerchen, Scientists under Hitler: Politics and the physics community in the
third Reich (New Haven, 1977), 90, 124, 170, 245.
O. Klein to W. Pauli, 8 March 1921, in A. Hermann, K. v. Meyenn and V. F. Weisskopf (eds),
Wissenschaftlicher Briefwechsel mit Bohr, Einstein, Heisenberg, u.a., i: 1919–1929 (Berlin,
1979), 27, cited in English translation by Miller, “A précis of Edmund Whittaker” (ref. 24),
95. For the case of Hilbert’s contribution, and the historiographical debate it has generated, see
John Stachel, “New lights on the Einstein–Hilbert priority question”, Journal of astrophysics
and astronomy, xx (1999), 91–101; L. Corry, J. Renn and J. Stachel, “Belated decision in the
Hilbert–Einstein controversy”, Science, cclxxviii (1997), 1270–3; F. Winterberg, “On ‘Belated
decision in the Hilbert–Einstein priority dispute’”, Zeitschrift für Naturforschung, lix/a (2004),
715–19; Tilman Sauer, “Einstein equations and Hilbert action: What is missing on page 8 of the
proofs for Hilbert’s ﬁrst communication on the foundations of physics?”, Archive for history of
exact sciences, lix (2005), 557–90.
W. Pauli, Theory of relativity, transl. by G. Field (New York, 1958), 2.
See M. N. Macrossan, “A note on relativity before Einstein”, The British journal for the philosophy
of science, xxxvii (1986), 232–4, p. 232; C. Kittel, “Larmor and the prehistory of the Lorentz
transformations”, American journal of physics, xlii (1974), 726–9, p. 726; see also W. Rindler,
“Einstein’s priority in recognizing time dilation physically”, American journal of physics, xxxviii
(1970), 1111–15. For a detailed analysis of Larmor’s works in relation to the reception of relativity
in England, see Warwick, op. cit. (ref. 11), chap. 7. For an internalist comparison of Larmor and
Lorentz, see O. Darrigol, “The electron theories of Larmor and Lorentz”, Historical studies in
the physical and biological sciences, xxiv (1994), 265–336.
Pauli, op. cit. (ref. 62), 3, our emphasis.
Ibid., our emphasis.
X. Léon in Bulletin de la Société Française de Philosophie, séance du 6 avril 1922, xvii (1922),
Charles Nordmann, “Un événement scientiﬁque — Einstein à Paris”, Le Matin, 23 March 1922, cited
by Michel Biezunski, Einstein à Paris (Saint-Denis, 1991), 16. One should recall that Einstein
had ﬁrst refused the invitation and then accepted it a few weeks later at the urging of the German
Foreign Minister, Walter Rathenau, who saw in it a gesture of reconciliation with the French.
The symbolic meaning of Einstein’s visit was clear to everyone.
Gustave le Bon, Compte rendus hebdomadaires de l’Académie des Sciences, séance du 6 juillet 1914,
clix (1914), 26–27. The note was communicated by the astronomer H. Deslandres.
Le Bon provides another example of the tendency to stake a claim to priority only after a theory has
been generally accepted. On the basis of his general philosophical ideas about the transmutation
of matter published in the years 1900 and collected in his book L’évolution de la matière (Paris,
1905), he claimed priority for the idea of transmutation after the Curies received the Nobel Prize
in 1903. On this episode, see Mary Joe Nye, “Gustave Le Bon’s black light: A study in physics
and philosophy in France at the turn of the century”, Historical studies in the physical sciences,
iv (1974), 163–95, pp. 191–2, and Benoit Marpeau, Gustave le Bon: Parcours d’un intellectuel,
1841–1931 (Paris, 2000), 250–63.
See the letters between Le Bon and Einstein in Albert Einstein, Oeuvres choisies 4. Correspondances
françaises, ed. by M. Biezunski (Paris, 1989), 186–99.
Einstein to Le Bon, 18 June 1922, ibid., 187.
Guillaume does mentions in a note that in their articles for the Encyklopädie der Mathematischen
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
Wissenschften, W. Pauli and M. Kottler do cite Poincaré adequately. As we have seen, Felix Klein
did make sure this was the case, at least for Pauli but maybe also for Kottler.
Hence Paul Kircherberger, La théorie de la relativité exposée sans mathématiques (Paris, 1922), was
translated from German with a preface by Max von Laue. Interestingly, the bibliography contains
some works by French authors (such as P. Langevin and J. Becquerel) but none by Poincaré.
Poincaré La mécanique nouvelle (ref. 33).
It has thus been published in German in 1910 as “Die neue Mechanik”, Himmel und Erde (Leipzig),
xxiii (1910), 97–116, and as a reprint: H. Poincaré, Die neue Mechanik (Leipzig and Berlin,
Guillaume, in Poincaré, “La mécanique nouvelle” (ref. 33), p. viii, italics in the original.
See Einstein to Jacob Laub, 20 March 1909, in CPAE, v, Document 143, 101. The paper on the
Maschinchen is in CPAE, ii, English translation, Document 48, 312–15.
See CPAE, iii, Documents 5 and 6.
Biezunski, op. cit. (ref. 67), 21–22 and, by the same author, “Einstein’s reception in Paris in 1922”
in Glick (ed.), The comparative reception of relativity (ref. 53), 169–88, pp. 176–7.
See the correspondence between them in vols viii–x of CPAE. For an analysis, see Angelo Genovesi,
Il carteggio tra Albert Einstein ed Edouard Guillaume:‘Tempo universale’ e teoria della relatività
ristretta nella ﬁlosoﬁa francese contemporanea (Milan, 2000); and Introduction to vol. x of
CPAE, pp. xliii–xlix.
CPAE, ix, English translation, Document 305, 255, and Document 330, 276.
For details see Biezunski, op. cit. (ref. 67), 47–48.
Paul Appell, Henri Poincaré (Paris, 1925), 86–87.
Louis de Broglie, Savants et découvertes (Paris, 1956), 51.
Discours du Général Dassault, in Le livre du centenaire de la naisssance de Henri Poincaré
1854–1954 (no author) (Paris, 1955), 101–2.
See the contribution of the civil engineer M. G. Darrieus, ibid., 136.
Discours de M. Le Président André Marie, ibid., 87.
Whittaker, op. cit. (ref. 18), 40. For a detailed critical analysis of Whittaker’s point of view, see
Gerald Holton, “On the origins of the special theory of relativity”, American journal of physics,
xxviii (1960), 627–36, reprinted in Holton, Thematic origins (ref. 6), 165–83; and Miller, op.
cit. (ref. 24).
P. C. Bridgman, review, Isis, xlvii (1956), 428–30, p. 429.
Born to Einstein, letter 102, 26 September 1953, in The Born–Einstein letters (ref. 56), 197.
Einstein to Born, letter 103, 12 October 1953, in The Born–Einstein letters (ref. 56), 199. It is in
this speciﬁc context that one should interpret what Einstein wrote two months later about the
upcoming 50th anniversary celebration of relativity in Berne. In a letter to A. Mercier, dated 9
November 1953, in which he declined for reasons of health to participate in this ceremony, he
wrote that he “hoped that one will also take care on that occasion to honor suitably the merits
of Poincaré and Lorentz” (cited and translated by Pais, op. cit. (ref. 14), 171). Having noted
the recent fuss generated by Whittaker about the proper attribution of the “discovery” it was
natural for him to mention that “due credit” should also be given to Lorentz and Poincaré. It
is doubtful however that he would have raised the question in the absence of Whittaker’s point
of view and if its impact on Born and other scientists had not put that question on the agenda.
Having become an icon and an old sage he could have forgotten that in 1920 he had answered
an American journalist about the history of special relativity, that it was Lorentz and himself
who had developed it (New York Times, 3 December 1920, cited by Pais, op. cit. (ref. 14), 171).
There was then no mention of Poincaré.
94. Max Born, review of Whittaker’s book, The British journal for the philosophy of science, v (1954),
261–3, p. 262.
95. Edmund Whittaker, “Albert Einstein, 1879–1955”, Biographical memoirs of Fellows of the Royal
Society, i (1955), 37–67, p. 42.
96. G. J. W[hitrow], “Obituary: Professor Sir Edmund Whittaker, F.R.S.”, The British journal for the
philosophy of science, vii (1956), 180–1, emphasis in the text.
97. P. G. Bergmann, “Fifty years of relativity”, Science, cxxiii (1956), 487–94, p. 488.
98. Born, Physics in my generation (ref. 59), 100–15, p. 106.
99. T. Kahan, “Sur les origines de la relativité restreinte”, Revue d’histoire des sciences et de leurs
applications, xii (1959), 159–65, p. 164.
100. A. Grünbaum, “The relevance of philosophy to the history of special relativity”, The journal of
philosophy, lix (1962), 561–74, pp. 573–4.
101. G. H. Keswani, “Origin and concept of relativity (Parts I and II): Reply to Professor Dingle and Mr
Levinson”, The British journal for the philosophy of science, xvii (1966), 149–52, p. 150; and G.
H. Keswani, “Origins and concept of relativity I”, The British journal for the philosophy of science,
xv (1965), 286–306; Part II, ibid., xvi (1966), 19–32; and Part III, ibid., xvi (1966), 273–94.
102. Herbert Dingle, “Note on Mr Keswani’s articles, Origins and concept of relativity”, The British
journal for the philosophy of science, xvi (1965), 244–6, pp. 244–5.
103. Karl Popper, “A note on the difference between the Lorentz-Fitzgerald Contraction and the Einstein
Contraction”, The British journal for the philosophy of science, xvi (1966), 332–3. The reply by
Keswani is in ibid., xvii (1967), 234–6.
104. Charles Scribner, Jr, “Henri Poincaré and the principle of relativity”, American journal of physics,
xxxii (1964), 672–8, p. 672.
105. H. M. Schwartz, “A note on Poincaré’s contribution to relativity”, American journal of physics, xxxiii
(1965), 170. He refers to the publication, in 1913, of the original papers of Lorentz, Einstein and
Minkowski, a collection that certainly contributed a lot to ﬁx the canonical history of relativity
and its main actors: Otto Blumenthal (ed.), Das Relativitätsprinzip (Leipzig, 1913), with notes
by Arnold Sommerfeld. E. B. Wilson reviewed it in Science, and characterized Einstein’s paper
as the “epochal formulation of the principle of relativity as a fundamental physical principle
independent of any hypothesis of shortening”, Science, xxxix (1914), 944.
106. H. M. Schwartz, “Poincaré’s Rendiconti paper on relativity”, Part I, American journal of physics,
xxxix (1971), 1287–94; Part II, ibid., xl (1972), 862–72; Part III, ibid., xl (1972), 1282–87; and
C. W. Kilmister (ed.), Special theory of relativity (New York, 1970), part of the series Selected
Readings in Physics.
107. Stanley Goldberg, “Henri Poincaré and Einstein’s theory of relativity”, American journal of physics,
xxxv (1967), 934–44; Camillo Cuvaj, “Henri Poincaré’s mathematical contributions to relativity
and the Poincaré stresses”, American journal of physics, xxxvi (1968), 1102–13; Camillo Cuvaj,
“Note on ‘Poincaré and relativity’”, American journal of physics, xxxviii (1970), 774–5; Arthur
I. Miller, “Comment on: Poincaré’s Rendiconti paper on relativity: Part I”, American journal of
physics, xl (1972), 923; Carlo Giannoni, “Einstein and the Lorentz-Poincaré theory of relativity”,
PSA: Proceedings of the biennial meeting of the Philosophy of Science Association, 1970, 575–89.
Curiously, despite its title, which refers explicitly to Whittaker’s thesis, this paper never discusses
Poincaré’s work nor cites his papers.
108. G. H. Keswani and C. W. Kilmister, “Intimations of relativity before Einstein”, The British journal
for the philosophy of science, xxxiv (1983), 343–54.
109. See for example Gary Gutting, “Einstein’s discovery of special relativity”, Philosophy of science,
xxxix (1972), 51–68; John Earman, Clark Glymour and Robert Rynasiewicz, “On writing the
THE COLLECTIVE CONSTRUCTION OF SCIENTIFIC MEMORY
history of special relativity”, PSA: Proceedings of the biennial meeting of the Philosophy of
Science Association, 1982, 403–16 ; and Kenneth F. Schaffner, comments on this paper in ibid.,
417–28. None of these papers cites Whittaker. For further examples of debates on who really
did what in relativity based on rational reconstructions, see the exchanges between E. Zahar and
Miller: Elie Zahar, “Einstein’s debt to Lorentz: A reply to Feyeraband and Miller”, The British
journal for the philosophy of science, xxix (1978), 49–60; Elie Zahar, Einstein’s revolution:
A study in heuristic (La Salle, 1989), chap. 5; and Miller, Albert Einstein’s special theory of
relativity (ref. 32). The fundamentally moral foundation of these enterprises is clearly indicated
by the frequent use of the word ‘injustice’ in relation to what was done to either Poincaré or
Einstein, depending on the author’s side in the debate. Already in 1960, Gerald Holton interpreted
Whittaker’s mistake in dating a paper of Lorentz from 1903 instead of 1904 as a sign that it was
“not merely a mistake” and revealed how the “biographer’s preconceptions” interacted with his
material; see Holton Thematic origins (ref. 6), 177. For an assessment of this literature see David
Cassidy, “Understanding the history of special relativity”, Historical studies in the physical
sciences, xvi (1986), 177–95. For more recent works, some of them generated by commemorative
events, see Arthur I. Miller, “Why did Poincaré not formulate special relativity in 1905”, in J.
L. Greffe, G. Heinzman and K. Lorenz (eds), Henri Poincaré: Science et philosophie (Nancy,
1994), 69–100; Olivier Darrigol, “Henri Poincaré’s criticism of ﬁn de siècle electrodynamics”,
Studies in history and philosophy of modern physics, xxvi (1995), 1–44; Shaul Katzir, “Poincaré’s
relativistic physics: Its origins and nature”, Physics in perspective, vii (2005), 268–92; Jean-Pierre
Provost and Christian Bracco, “La relativité de Poincaré en 1905 et les transformations actives”,
Archive for history of exact sciences, lx (2006), 337–51; and Cerf, op. cit. (ref. 2).
Jean Dieudonné, “Poincaré, Jules Henri”, in Dictionary of scientiﬁc biography, xi, 51–61, p. 59.
Keswani, “Origins and concept of relativity (II)” (ref. 101).
Herbert E. Ives, “Revisions of the Lorentz transformations”, Proceedings of the American
Philosophical Society, xcv (1951), 125–31.
On Whittaker’s opposition to relativity see Jeffrey Crelinstein, Einstein’s jury: The race to test
relativity (New Jersey, 2006), 31–35.
For reasons of completeness of our survey, this section and the next summarize an analysis developed
in more details in Gingras, op. cit. (ref. 2).
Gérard Pilé, “Henri Poincaré (1873) et la relativité: Avant-propos”, La jaune et la rouge, April
Jules Leveugle, “Poincaré et la relativité”, La jaune et la rouge, April 1994, 30–51, pp. 49–50.
On Ives, see Herbert E. Ives The Einstein myth and the Ives papers (Old Greenwich, 1980).
Jules Leveugle, La relativité, Poincaré et Einstein, Planck, Hilbert: Histoire véridique de la théorie
de la relativité (Paris, 2004).
Respectively in Le Nouvel Observateur, 5 August 2004, 51, and Le Monde, 15 April 2005, 8.
Philippe Thomine (director), Tout est relatif, Monsieur Poincaré!, produced by Vidéoscop, 2005.
The ultimate national icon of France is of course Descartes. See François Azouvi, Descartes et
la France (Paris, 2002), which provides a fascinating cultural history of how the philosopher
came to incarnate France itself.
For a detailed analysis of the rhetoric of the ﬁlm see Gingras, op. cit. (ref. 2). On ‘counter-memory’,
see Nathalie Zemon Davies and Randolph Starn, “Introduction” to the Special Issue on “Memory
and counter-memory”, Representations, xxvi (1989), 1–6.
Christopher Flood, “The politics of counter-memory on the French extreme right”, Journal of
European studies, xxxv (2005), 221–36, pp. 222 and 234.
François Delaporte (ed.), A vital rationalist: Selected writings from Georges Canguilhem (New
York, 1994), 51.
124. J. T. Clark, “The philosophy of science and the history of science”, in Marshall Clagett (ed.),
Critical problems in the history of science (Madison, 1959), 103–40, p. 103. Interestingly, in
his critique of Whittaker, Grünbaum also quotes this author to discredit the method used by the
mathematician-historian. See Grünbaum, op. cit. (ref. 100), 573.
125. Delaporte, op. cit. (ref. 123), 51. Koyré is cited by Canguilhem on the same page.
126. For a complementary discussion see “Introduction” of Loren Graham, Wolf Lepenies and Peter
Weingart (eds), Functions and uses of disciplinary histories (Dordrecht, 1983), pp. ix–xx.
127. Peter Galison, “Introduction”, Isis, xcv (2004), 610–13, p. 611.
128. Darrigol, “The electrodynamic origin of relativity” (ref. 3), 311.
129. Ibid., 311.
130. For a general discussion of attribution of a ‘discovery’, see A. Brannigan, The social basis of scientiﬁc
discoveries (Cambridge, 1981).
131. On the autonomy of historians of science, see Yves Gingras, “The search for autonomy in history of
science”, in Jurgen Renn et Kostas Gavroglu (eds), Positioning the history of science (Dordrecht,