An Essay On Thomas Kuhn's First Scientific Revolution

An Essay on Thomas Kuhn’s
First Scientific Revolution,

The Copernican Revolution
N. M. SWERDLOW
The University of Chicago
T HE C OPERNICAN R EVOLU TION , Planetary Astronomy in the

Development of Western Thought, to give its complete title,
Thomas Kuhn’s first book, may be the second best selling book
ever written on the history of science. As I write this essay, the Harvard
University Press edition is in its nineteenth printing, and that does not
include its many years as a Vintage Book. It was also one of Kuhn’s
first publications in the history of science; previously he had published
but six papers, on seventeenth-century chemistry and on the Carnot
cycle, five of which were short notes. The book had its origin in lec-
tures for a course at Harvard using a historical approach to teach, not
so much science as an understanding of science, to students outside the
sciences, that were later rewritten for a more general readership. This
history is important to understanding the character of the book, and
even the history has a history, which is worth telling.
On 16–18 September of 1936 Harvard University observed its Ter-
centenary with a very grand commemoration at a very difficult time. It
was, as one always says, “the Depths of the Depression,” the Tercente-
nary Fund had not done well, and the alumni, who were counted on
for contributions, were mostly of the sort who referred to the guest of
This paper was originally presented at the conference on The Legacy of Thomas S. Kuhn
at the Dibner Institute for the History of Science and Technology in November 1997. I would
like to thank I. Bernard Cohen, Anthony Grafton, John Heilbron, and Karl Hufbauer for
very helpful information and advice in its revision; Hufbauer’s paper, “Thomas Kuhn’s
Discovery of History (1940–1958),” has also been invaluable.
Abbreviations:
CR Kuhn, The Copernican Revolution
ET Kuhn, The Essential Tension
SR Kuhn, The Structure of Scientific Revolutions
US Conant, On Understanding Science
PROCEEDINGS OF THE AMERICAN PHILOSOPHICAL SOCIETY VOL. 148, NO. 1, MARCH 2004
[ 64 ]
thomas kuhn’s first scientific revolution 65
honor, the honorary chairman of the United States Harvard Tercentenary

Committee, Harvard’s most famous alumnus, President Roosevelt
(’04), as “that man in the White House” and “a traitor to his class.”
Few despised Roosevelt more than Harvard’s former president, Abbot
Lawrence Lowell, who feared that he would use the occasion “to make
a stump speech” for the November election, and at first refused to pre-
side over the meeting of the alumni at which Roosevelt was to speak.
But despite these forebodings, it turned out to be a splendid occasion,
featuring a floating concert on the Charles by the university band accom-
panied by fireworks for a crowd estimated at 350,000 on both sides of
the river. Seventy-one papers presented at a conference preceding the
festivities, reported at length in newspapers, were followed by no less
than sixty honorary degrees awarded to the likes of Compton, Edding-
ton, R. A. Fisher, G. H. Hardy, Levi-Civita, Malinowski, Piaget, and,
for reasons that escape me, C. G. Jung. The principal events of the last
day, the Tercentenary itself, were, in the morning, the Salutatory Oration
in Latin by Professor Edward K. Rand (Salvete Omnes! it begins), an
address by Governor James M. Curley, in no way an alumnus, who
referred to Roosevelt as “dear to the hearts of the alumni of Harvard,”
the Oration by Harvard’s president, James Bryant Conant, and the award-
ing of honorary degrees; in the afternoon, at the meeting of the Harvard
Alumni Association, speeches by Learned Hand (’93), President James
Rowland Angell of Yale, who warned of the dangers to endowed institu-
tions of property, income, and estate taxes, and President Roosevelt.
The ceremonies, for an audience of seventeen thousand, were held
in Harvard Yard between Widener Library and Memorial Church, and
mostly it poured rain—there is a picture of Roosevelt, who refused an
umbrella, sitting defiantly in a wet morning coat and top hat—leading
President Angell to remark in his speech that he had heard a saturated
graduate say, “This is evidently Conant’s method of soaking the rich.”
Due to the rain, the afternoon session was adjourned to Sanders Theatre,
where Lowell gave Roosevelt the most perfunctory introduction ever
given to a president of the United States outside of Congress, “The next
speaker it would be impertinent for me to introduce to you. . . . Gentle-
men, the President of the United States!” The opening of Roosevelt’s
speech is also memorable. As Conant later reported, he began:
A hundred years ago, when Harvard was celebrating its two hun-
dredth anniversary, Andrew Jackson was President and Harvard men
were sore afraid. Fifty years ago, when Harvard was celebrating its
two hundred and fiftieth anniversary, Grover Cleveland was President
and Harvard men were sore afraid. Today, I am President. . . .
Roosevelt stopped, even the crotchety alumni applauded, and the rest
of the speech, to as hostile an audience as could be imagined, went
66 n. m. swerdlow
splendidly. He later wrote to Felix Frankfurter, “I told the boys afterwards

that I had stuck out my chin and said ‘hit me’—and nobody dared.” 1
A high point of the festivities, and it was a triumph, was Conant’s
President’s Oration, or Tercentenary Address, as he later called it,
which brought him national fame—he was soon on the cover of Time
and profiled in The New Yorker—and which he thought enough of
more than thirty years later to reprint in his autobiography. In a way, it
is a speech of the sort usually given at academic ceremonies, particu-
larly anniversaries, on “what is a university?” but the times gave the
issue particular urgency. He touched upon various subjects, among
them freedom of inquiry and dissent in the search for the truth, but
also an increasing demand to turn universities into schools for profes-
sional, vocational training, a euphemism for students’ pursuing education
for profitable careers and universities’ pursuing students for profitable
tuitions, something that could not possibly happen in our time. “A
wave of anti-intellectualism is passing round the world,” he said, “. . . a
protest against the benefactions of the learned world, . . . a weariness
at the ever-increasing wealth of new knowledge poured at our feet by
scholars of the arts and letters no less than by the scientists.” His solu-
tion lay in a reform of liberal education. The older education, based
upon the classics and mathematics, which provided a common back-
ground and steadied the thinking of all educated men, is gone and can-
not be brought back, and we must find its modern equivalent. This he
saw in historical studies. “In my opinion, it is primarily the past devel-
opment of our modern era which we must study, and study most
exhaustively and critically. We must examine the immediate origins of
our political, economic and cultural life, and then work backwards. . . .
For the development of a national culture based on a study of the past,
one condition is essential. This is absolute freedom of discussion, abso-
lute unmolested inquiry.”2
1 On the Harvard Tercentenary, I have followed Conant, My Several Lives, 146–56,
Greene, passim, Freedman, 322–30, 355–56, which Conant thought indiscreet, and
Hershberg, 90–92, 97–99. The volume by Professor Jerome D. Greene, the director of the
Tercentenary, runs to no less than 492 pages. Conant, who evidently relied on his memory,
improved upon Roosevelt’s remarks; Greene and Freedman give a longer but less sharp
version, although the point is the same, and print Roosevelt’s entire address, prepared by
Frankfurter, which was not a “stump speech” and contains a (long and obscure) quotation
from Euripides. (The one comparable event I can think of was when President Kennedy,
invited by Yale to receive an honorary degree on 11 June 1962, not exactly to the delight of
the Old Blues, opened a speech at Commencement, which in principle does not allow
speeches, with the words, “It might be said now that I have the best of both worlds, a
Harvard education and a Yale degree.” The audience applauded and the president proceeded
to give a fairly technical lecture on government monetary and fiscal policy.)
2 Conant, My Several Lives, 651–58.
Conant later remarked that his defense of free inquiry was widely
applauded while his suggestion for historical studies found few sup-
porters either in 1936 or thereafter. In that year Harvard had estab-
lished a Committee on Higher Degrees in the History of Science and
Learning, the predecessor of the history of science department, but this
was a graduate program not concerned with general education. It was
to be ten years before Conant could put his ideas into practice. In 1943
he appointed a committee under the dean of the Faculty of Arts and
Sciences, Paul H. Buck, to write a report on “The Aims of General
Education in a Free Society,” the project being supported with a grant
of $60,000. Conant twice appeared before the committee, and sug-
gested a course on “The Advancement of Knowledge in Modern Times,”
to show how knowledge had been advanced in the last four hundred
years, choosing the examples to correspond to certain decisive periods
in the development of various disciplines. What he had most in mind
was the history of the experimental sciences. In 1945 the committee
presented its report, General Education in a Free Society, known as the
“Redbook,” a comprehensive examination of secondary and collegiate
education, one chapter of which concerned general education at Har-
vard. Now, if there is a rule that governs innovation in general educa-
tion it is this: Programs for the reform of general education are either
unrealistic or inconsequential. The University of Chicago, which could
hardly be accused of vocational education, chose the former in 1942
and has still not completely recovered. The Harvard Plan falls into the
latter category, which is probably just as well, and consisted essentially
of requiring students to take at least three general education courses,
one each in the humanities, social sciences, and natural sciences, as
well as some other distribution requirements. Compared to the earlier
change from electives under President Charles William Eliot to concen-
tration and distribution requirements under Lowell, this was hardly
radical, although the creation of separate courses for general educa-
tion, which was not in itself new and which most universities continue
to this day, has had lasting effects in separating general and specialized
education. Presumably through Conant’s influence the natural science
courses were to be given a significant, even central, historical compo-
nent. The curious justification offered by the Redbook was that, as
much of elementary science is no longer challenging to scientists, or
even to well-informed students, a historical approach might make it
more interesting by showing that topics that are now “scientific detri-
tus” were once matters of concern and controversy.3
3 General Education in a Free Society, 225–26; I am not making this up. Nevertheless, the
Redbook remains a pertinent and thoughtful evaluation of education in America, at once

68 n. m. swerdlow
Conant described the new method of instruction in three Terry Lec-

tures presented at Yale in the fall of 1946, which were published the
following year as On Understanding Science, and then he practiced
what he preached in the fall of 1947 by offering one of the first courses
of what became Natural Sciences 4 (originally 11a), “The Growth of
the Experimental Sciences.” Bernard Cohen helped in assembling the
historical materials and preparing the manuscript of the lectures for
publication, and the assistant for the course was Thomas Kuhn, who
gave several lectures on seventeenth-century mechanics. By “under-
standing science” Conant meant approaching science with the special
point of view of one who has actually done it, something that a scien-
tist has and a layman does not, not because of a lack of scientific
knowledge, but because, through lack of experience, “he has no ‘feel’
for the Tactics and Strategy of Science.” It was the Tactics and Strategy
of Science, not science itself, that Conant believed essential to every edu-
cated person in a free and democratic society, certainly to those gradu-
ates of Harvard who were one day to be “lawyers, writers, teachers,
politicians, public servants, and businessmen.”4 This may be accom-
plished by approaching science either historically or analytically, that
is, philosophically, but Conant considered the historical method far
better in giving a picture of the true difficulties of science, which the
philosophy of science, as exemplified by Karl Pearson’s The Grammar
of Science, simplified into a handful of rules that no one followed any-
way. “The stumbling way in which even the ablest of the early scientists
had to fight through the thickets of erroneous observations, misleading
generalizations, inadequate formulations, and unconscious prejudice is
the story, which it seems to me needs telling.” This could best be done
through study of “case histories,” similar to the case method already
used in law and business schools, as it is to this day, and the use of case
histories was not to teach the history of science, but as the best substitute
traditional and liberal, perhaps insufficiently concerned with the effects of dire poverty—the
depression is seldom mentioned, segregation and discrimination never—yet calling upon
the federal government to enlarge and more nearly equalize educational resources and
opportunity throughout the nation. It calls training in foreign languages in secondary schools
a “Copernican step” (pp. 119–27). Conant’s account of the report and the natural science
courses is in Conant, My Several Lives, 363–73. Kuhn published a comment on the report in
the Harvard Alumni Bulletin (22 Sept. 1945), 23–30, which he later listed as the first publi-
cation on his curriculum vitae.
4 US, 1. He wrote to J. W. Shirley, “. . . I have in mind starting a young man who is not at
all interested in science down the road which will end with his being a citizen—we hope a
leading citizen—who can read about modern science and talk to modern scientists with some
kind of understanding.” Shirley, 421. Conant wrote a great deal about education in general
and scientific education in particular, and what he wrote was of considerable importance in
American education following the war. But it goes beyond the subject of this essay.
for actually engaging in scientific research. Although Conant’s purpose

was to show how science is practiced at the present day, it seemed
advisable to choose the case histories from earlier periods, both because
less scientific and mathematical background is required of the student,
and because one can more clearly see “the necessary fumblings of even
intellectual giants when they are also pioneers” and understand “how
difficult it is in fact to carry out glib scientific precepts.” Clearly, Con-
ant greatly preferred history as philosophy teaching by example to phi-
losophy itself, which he regarded as static, while history, like science, is
dynamic.5
I should hope to illustrate the intricate interplay between experiment,
or observation, and the development of new concepts and new gener-
alizations; in short, how new concepts evolve from experiments, how
one conceptual scheme for a time is adequate and then is modified or
displaced. I should want also to illustrate the interconnection between
science and society about which so much has been said in recent years
by our Marxist friends.6
Conant considered the sciences to be “accumulative knowledge” in
which one can assuredly state that advances have been made, that
there has been progress, unlike philosophy, poetry, and the fine arts in
which it is by no means clear that the present represents either advance
or regress compared to the past. If Galileo or Newton could observe
the present, there is no doubt of their judgment; for Locke or Milton or
Michelangelo a judgment of progress is hardly certain. It is in fact this
view of science as accumulative knowledge that stands behind the
value of the historical approach to understanding it, and in fact stands
behind Conant’s answer to the question “What is Science?” which he
says he had originally intended to dodge but decided he could answer,
“not in analytical but in historic terms.” His answer is so interesting,
and so important to Kuhn’s Copernican Revolution, that I shall quote
it in full in two paragraphs, as I do not want to simplify or misrepre-
sent anything by paraphrase.
As regards a first approximation, we may say that science emerges
from the other progressive activities of man to the extent that new
concepts arise from experiments and observations, and the new con-
cepts in turn lead to further experiments and observations. The case
histories drawn from the last three hundred years show examples of
fruitful and fruitless concepts. The texture of modern science is the
result of the interweaving of the fruitful concepts. The test of a new
5 US, 11–20.
6 US, 18.
70 n. m. swerdlow
idea is therefore not only its success in correlating the then-known

facts but much more its success or failure in stimulating further exper-
imentation or observation which in turn is fruitful. This dynamic
quality of science viewed not as a practical undertaking but as devel-
opment of conceptual schemes seems to me to be close to the heart of
the best definition. It is this quality which can be demonstrated only by
the historical approach, or else learned by direct professional experience.
To illustrate what I have in mind let us imagine a period in the
future when all interest in scientific investigation had ceased but
the relatively simple conceptual schemes about matter and energy, the
solar system and the basic facts of chemistry of the late nineteenth
century were accepted and widely taught. Would the people of that
day “understand” science as the late Victorians did? Not to my mind.
There would be little difference in their intellectual outlook from that
of a people who accept their cosmology as part of a revealed religion.
If this be so, the characteristic of the scientific age in which we live lies
not in the relative adequacies of our conceptual schemes as to the uni-
verse but in the dynamic character of these concepts as interpreted by
both professional scientists and laymen. Almost by definition, I would
say, science moves ahead.7
Note that Conant’s definition is “dynamic,” a sort of scientific selec-
tion in which a “concept,” which arose from experiments and observa-
tions, survives, propagates, is “fruitful,” not just in accounting for what
is known, but far more in stimulating further experiments and observa-
tions that are “fruitful,” that is, lead to more concepts. Were “conceptual
schemes” believed and taught, but not applied to scientific investigation,
there would be no “understanding” of science, only something like
religious faith. We live in a scientific age not in virtue of what we know,
but what we do. Science must advance to stay alive, must advance to
meet the definition of science. This is a very sophisticated definition,
really of modern science of, say, the last four hundred years, not a mere
invocation of progress; indeed, it admits that concepts lead to dead
ends at least as often as advancement, which can hardly be continuous,
for it is a selection of what is “fruitful” that determines what survives,
which can happen only so long as scientific research is active. And it is
the definition of a scientist with an eye for history, not of a philosopher
concerned with logical analysis or method. The words “concept,” “con-
ceptual scheme,” and “fruitful” are terms of art that are used in The
Copernican Revolution, where we shall consider their meaning.
To illustrate the Tactics and Strategy of Science, and possibly also
his definition of science, Conant presents three case histories, the first
7 US, 24–25.
on the elasticity of the air, concentrating on Boyle, the second, very

briefly, on Galvanic electricity, and the third on phlogiston and the dis-
covery of oxygen, concentrating on Lavoisier, drawing from each some
Tactics and Strategies of Science that are summarized at the end. Here
are some examples concerning the origin of new concepts that will be
seen to be pertinent to The Copernican Revolution:
New concepts may result from a consideration of difficulties inherent
in an old concept, the conflict between the conceptual scheme and
certain stubborn facts. These facts may be well known but not yet
explained.
New concepts may evolve through a series of successive approxi-
mations from older ones, the modification never being so drastic as to
constitute a complete jettisoning of the older idea.
A new concept may be revolutionary and after its formulation a
host of old facts may be fitted into the new scheme and many new
facts discovered.
Erroneous observations or interpretations of experiments fre-
quently persist and confuse the development of new concepts.8
On Understanding Science was quite specifically a description of a ped-
agogical method. After three years of teaching by the case history
method, Conant wrote a greatly expanded version, Science and Com-
mon Sense (1951), to convey to the general reader what was conveyed
to the students in the course. Notable throughout is his skepticism
about a prescriptive scientific method, a belief that the only way to
learn about science is from science, with and, in particular, because of
its uncertainties, and a great number of examples from many fields to
back up his points. It is a thoughtful book, a challenging book, also an
idiosyncratic book. I do not know what success it had in its day.
Conant’s were the first of the Case Histories prepared for the natu-
ral science course over the next few years by Conant, Leonard K.
Nash, Duane Roller, and Duane H. D. Roller that were published sepa-
rately and eventually collected in the two-volume Harvard Case Histo-
ries in Experimental Science (1957). They consist of excerpts from
original papers with extensive, sometimes very extensive, introductions
and expositions by the editors, and in addition the students were given
lists of supplemental readings in philosophy, history, and the social sci-
ences. Conant had acknowledged the importance of intellectual along
with economic, social, and technological factors in the development of
science, and even envisioned a course on “The Growth of Science as
an Organized Social Activity,”9 but aside from the occasional nod to
8 US, 101–04.
9 US, 106.
72 n. m. swerdlow
technology, the Case Histories are confined to pure science, perhaps as

a result of the restriction to experimental science in which the role of
external, intellectual factors is less apparent than in the theoretical sci-
ences. Philosophy is not supposed to intrude into the laboratory, at
least, if it is well run. Apparently the students did not carry out experi-
ments of any kind, which seems a pity when studying the experimental
sciences. The subjects covered include Boyle’s experiments in pneumat-
ics, Priestley and Lavoisier on phlogiston and the discovery of oxygen,
Pasteur and Tyndall on fermentation, Pasteur and Tyndall on sponta-
neous generation, all by Conant; Dalton, Gay-Lussac, and Avogadro
on the atomic-molecular theory, the discovery of the carbon-oxygen
cycle in plants in the late eighteenth century, by Nash; the rise and
decline of the caloric theory, concentrating upon Black, Rumford, and
Davy, by Duane Roller; and the concept of electrical charge, beginning
in antiquity but mostly in the eighteenth century to Coulomb, by
Duane Roller and Duane H. D. Roller. Most, although not all, of them
are on the overthrow of something old and the discovery of something
new, as phlogiston and oxygen, on the replacement of one “conceptual
scheme” by another, and Conant’s terminology turns up frequently.
It is instructive to read these Case Histories today, which I suppose
no one does, but I have at least in part. How effective they were in
teaching the Tactics and Strategy of Science to the undergraduates I do
not know although there are many who could still be asked. Conant
published a progress report containing a description and assessment of
the course as early as 1949, when four of the Case Histories were in
use, admitting that the response of the students was mixed, and by
1959 there was so much criticism of the whole historical method that
students were permitted to fill the natural science requirement by tak-
ing courses from science departments. It appears as though there was
always some conflict or misunderstanding between Conant’s Tactics
and Strategies of Science and what others, whether students or outside
critics, may have expected of the course.10 With all of modern science
to learn, one must wonder how interested students could ever be in
the evolution of oxygen from mercurius calcinatus per se, however
10 Conant, The Growth of the Experimental Sciences , 5–10; Shirley, 422–23, criticized the
course for attempting to do too much, integrate scientific, social, and humanistic studies, but
actually doing too little, the experimental sciences narrowly conceived; Keller, 17, on the
later criticism. My own opinion, for what it is worth, is that the history of science is not an
effective way to teach science, or about science, in general education, and one is better off
with science courses. But I also believe that students in the sciences can benefit from serious
technical, historical study of their own subjects—it is a part of knowing the subject
thoroughly and it is better to get these things right than wrong—and that serious history of
science properly belongs within science departments.
important that discovery may have been in the hands of Priestley and
Lavoisier.
Nevertheless, as history of science itself, which admittedly was not
their purpose, the Case Histories are pretty good. The first thing one
notices is that what was read by Harvard undergraduates in a general
education course fifty years ago contains more science, and on a higher
technical level, than most of the reading assigned to graduate students
in history of science courses today, higher even than the scientific level
of most scholarly books and articles written on these subjects, or in
general, by historians of science in the last twenty or thirty years. And
problem sets included in some of the Case Histories, as well as specimen
examinations published by Conant in his progress report, would be over
the heads, not only of most of today’s graduate students in the history of
science, but most of their professors. The reason is not far to seek. The
Case Histories were written and the natural science courses were taught
by scientists, who, in Conant’s words, had a “feel” for the Tactics and
Strategy of Science, knew a good deal of science, and found science
interesting, the very opposite of the conditions that now prevail among
historians of science. This is a sensitive subject with a history of its own,
and I hesitate to say more about it here, but it is worthy of reflection.
The published Case Histories were all in the experimental sciences,
but to judge by Conant’s progress report, the more theoretical sciences
were also treated, bringing us closer to our principal subject. 11 In 1947
Kuhn had lectured in Conant’s course on the origins of seventeenth-
century mechanics, which was notable because, as he later wrote, it
was in preparing the lectures during the summer that he discovered
that Aristotelian physics was not so absurd as it first appeared if only
one understood what was meant by words like “motion,” “change,”
and “quality” that differ greatly from what we mean by them today. “I
did not become an Aristotelian physicist as a result, but I had to some
extent learned to think like one.”12 The next time he participated in the
course, in the fall of 1950, he began lecturing on the subjects of The
11 What follows is brief. A more detailed account of Kuhn’s teaching in the natural
sciences course and other courses in the history of science can be found in Hufbauer’s paper
and in Kuhn’s own reminiscences in “A Discussion with Thomas S. Kuhn.” Hufbauer notes
that the natural science course was successively renamed “Research Patterns in the Physical
Sciences” and “The Process of Research in the Physical Sciences,” which show an extension
beyond the original experimental sciences. According to the listings in the 1951–52 Harvard
University Courses of Instruction sent to me by Hufbauer, there were then five natural science
courses: 1. The Physical Sciences in a Technical Civilization; 2. Principles of Physical Science;
3. The Nature and Growth of the Physical Sciences; 4. Research Patterns in Physical Science; 5.
Principles of Biological Science.
12 ET, xii. Readings for the course among Kuhn’s papers at MIT include Aristotle,
Benedetti, and Galileo. I am again grateful to Karl Hufbauer for advising me to consult
Kuhn’s papers following the presentation of the original version of this paper.
74 n. m. swerdlow
Copernican Revolution, which he continued to do each year before he

received a Guggenheim in 1954–55, during which he wrote most of the
book, and then departed for Berkeley in 1956 where the book was
completed. In the fall 1953 and 1955 he taught a history of science
course on the “Rise of Scientific Law: Aristotle to Newton,” also related
to the book. The original lectures contain the principal parts of the
book, both scientific and historical, but the book itself was enlarged to
go beyond the natural sciences course and address the general reader. 13
Still, vestiges of the course remain, not only in the considerable techni-
cal content, including a “Technical Appendix,” all of it explained with
the greatest clarity for which the book remains exceptional, but in the
description of its original goal, to teach not science itself but something
like the Tactics and Strategy of Science, which Kuhn put in his own,
perhaps prophetic, words: “Since students in this General Education
course do not intend to continue the study of science, the technical facts
and theories that they learn function principally as paradigms rather
than as intrinsically useful bits of information.”14
The Copernican Revolution may have begun as something like the
Case Histories, but became far more original and far more important,
which is why it is still read and the Case Histories are not. The Case
Histories were mostly concerned with changes of “conceptual schemes,”
some momentous, some not, and this was also Kuhn’s concern, although
on a much larger scale and differing from the case histories in two ways.
The first was the consideration of intellectual as well as scientific his-
tory, to which Kuhn drew attention in his preface, pointing out that the
Copernican Revolution had a scientific and intellectual component, both
in its origin and its consequences, and that the most important novelty
of his study was treating both as equally fundamental. 15 The other dis-
13 “. . . I started with lectures on the Copernican revolution. The book really, though it’s
got more detail, was modeled very precisely [on those lectures]; it’s an extended case history
of the sort there.” “A Discussion with Thomas S. Kuhn,” 172. Kuhn’s papers confirm the
accuracy of this description, lectures and readings on a large case history, including eighteen
problems in astronomy; nevertheless, a great deal of expansion and rewriting went into the
book.
14 CR, ix.
15 CR, vii–viii. This is also true of Herbert Butterfield’s The Origins of Modern Science,
1300–1800 (London, 1949, rev. ed. London, 1957), which Kuhn noted “had particular
influence on the structure of this book” (CR, 283). I assume this refers both to the inclusion
of intellectual history and to Butterfield’s emphasis on “those cases in which men not only
solved a problem but had to alter their mentality in the process, or at least discovered
afterwards that the solution involved a change in their mental approach” (Collier ed., p. 8).
Kuhn was already concerned with this independently. For two decades following its publication
Butterfield’s excellent book was of the greatest importance in creating general interest in the
history of science, and was approached or equaled in this only by CR. I would guess that
prior to SR they were the two most widely read books on the subject.
tinction from the Case Histories was not enunciated in the preface, but
was very much on Kuhn’s mind and is central to the book. The Case
Histories were concerned with change in science, with the replacement
of one conceptual scheme by another, but none addressed the question
of how it takes place in general or in all its specific complexity. Conant
had made a start at this in his section on the definition of science in On
Understanding Science, and he clearly recognized in his foreword to
The Copernican Revolution, which is worth reading, that Kuhn had
gone much farther.16 The subject had already concerned Kuhn for sev-
eral years. “After stumbling upon the concept of a scientific revolution
in 1947,” by which he meant the change from Aristotelian to Galilean
and Newtonian mechanics—“While discovering history, I had discov-
ered my first scientific revolution”—he turned his attention to the ques-
tion of how this happens, along with the history of science in general.
In 1951 he presented his ideas on the subject in a series of Lowell Lec-
tures, The Quest for Physical Theory: Problems in the Methodology of
Scientific Research, “but the primary result of that venture was to con-
vince me that I did not yet know enough history or enough about my
ideas to proceed toward publication.”17 The Copernican Revolution,
Kuhn’s first published attempt at an answer, may be understood as a
great case history of one of the monumental changes in the history of
science in order to provide an explanation of how so great a revolution
happens, and I use the present tense deliberately. In this sense, it is his
first scientific revolution. Kuhn regarded the book in later years as
something he had written for a course that was not really in his field—
as he put it to me, he was, after all, a physicist, and what he knew best
was modern physics—in which he was still finding his way in the history
16 Among Conant’s remarks (CR, xviii): “This book is no superficial account of the work
of scientists; rather it is a thorough exposition of one phase of scientific work, from which
the careful reader may learn about the curious interplay of hypothesis and experiment (or
astronomical observation) which is the essence of modern science but largely unknown to the
nonscientist. . . . I wish to register my conviction that the approach to science presented in
this book is the approach needed to enable the scientific tradition to take its place alongside
the literary tradition in the culture of the United States.”
Kuhn’s acknowledgment of Conant in the preface (CR, ix): “Many friends and colleagues,
by their advice and criticism, have helped to shape this book, but none has left so large or
significant a mark as Ambassador James B. Conant. Work with him first persuaded me that
historical study could yield a new sort of understanding of the structure and function of
scientific research. Without my own Copernican revolution, which he fathered, neither this
book nor my other essays in the history of science would have been written.
“Mr. Conant also read the manuscript, and its early chapters show many signs of his
productive criticism.”
According to Hershberg, 410, Conant was critical of SR.
17 ET, xiii, xvi. The lectures are among Kuhn’s papers. See also Hufbauer’s paper.
76 n. m. swerdlow
of science. Nevertheless, it is the work of an extraordinarily creative

and intelligent man, as creative and intelligent as anyone I have ever
known, taking on an extraordinarily large and important subject; and
even if he would have done it differently later, it will always be of inter-
est both for itself and for understanding Thomas Kuhn himself. In
what follows, I shall go through the book selectively, sometimes para-
phrasing or summarizing, sometimes quoting directly, both with and
without quotation marks, more or less following its general argument,
but concentrating on those sections in which Kuhn’s exposition is con-
cerned with the causes and consequences of conceptual change in science,
in this way using The Copernican Revolution to show the development
of his own ideas. Since I also have some independent interest in the
subject of the book, which is why I chose to write this essay, I shall also
expand or comment upon its exposition in places.
Kuhn begins by asking the question “What does the phrase ‘Coper-
nican Revolution’ mean?”18 His answer defines the great breadth of his
study, going well beyond what is usually understood by its title. It
means first a reform in the fundamental concepts of astronomy, second
a radical alteration in other sciences necessitated by the motion of the
earth, and third a still more comprehensive effect on philosophy, reli-
gion, and values resulting from no longer seeing the earth as the unique
center of God’s creation. The Copernican Revolution broadly under-
stood has all of these meanings, astronomical, scientific, and philosoph-
ical, and must therefore be studied as both scientific and intellectual
history. Let me add for specificity that by “Copernican Revolution”
Kuhn means not just the change from the Ptolemaic to the Copernican
planetary system, although that is a central part of the story from
which it takes its name, but the change from Aristotelian to Newtonian
cosmology, which also seems to be seldom understood. It is in this
sense that it can stand beside the change from Aristotelian to Newto-
nian mechanics as a monumental revolution, and the two revolutions
are of course closely, indeed, causally and reciprocally, related. That is
what I mean when I say it is a study of great breadth, for it tells one of
the two or three largest stories in the history of science, and perhaps
the largest because the story goes beyond the history of science to the
resulting transformations in philosophy, in religion, in world-view, to
use a term usually applied incorrectly but in this case correctly since a
cosmology is more than a physical description of the universe. Further,
Kuhn remarks, a study of the Copernican Revolution will show in what
way scientific concepts, although believed for the best of reasons, are
18 CR, 1–4.
unlikely to prove final. Not that the mutability of its fundamental con-
cepts is an argument for rejecting science. “Each new scientific theory
preserves a hard core of the knowledge provided by its predecessor and
adds to it. Science progresses by replacing old theories with new,” and
the history of Copernican theory, as of any scientific theory, can illus-
trate the processes by which scientific concepts evolve and replace their
predecessors.
The first subject considered is pre-scientific cosmology, as that of
Egypt, which is actually quite important for already here the point is
made that from the beginning a cosmology must both explain observed
phenomena and provide a “psychologically satisfying world-view,” two
functions that are not always compatible, for the astronomer, entirely
for astronomical reasons, may destroy the very cosmology that made
the universe meaningful. Right here is a good part of the explanation
of the resistance to Copernican theory, and we see why such resistance
is hardly something new. Ancient scientific cosmology is explained phe-
nomenologically, through first describing the characteristic apparent
motions of the stars and the sun. The observations of phenomena are
objective, but the theories or “conceptual schemes” derived from them
depend upon the imagination of scientists and “are subjective through
and through.” The ancient cosmology developed at length is the Greek,
the basis of which is the “two-sphere universe,” a small, spherical
earth located at the center of a far larger sphere carrying the stars; if
the radius of the larger sphere is taken to be indefinite, this is the model
now called the “celestial sphere.” It is not itself a cosmology, for as yet
it lacks the moving bodies, the planets, but a framework for a cosmol-
ogy. The two-sphere universe can itself account for the phenomena of
spherical astronomy, that is, the risings and settings with respect to the
local horizon of all bodies considered as points on the celestial sphere,
including the apparent motion of the sun projected onto the celestial
sphere as the ecliptic.
Kuhn calls the two-sphere universe a “conceptual scheme,” a term
we have already used many times without defining, and we have noted
that it was used by Conant, who sometimes glossed it as “theory,” as
Kuhn does also, although it means more than that, as he goes on to
explain.19 A conceptual scheme is a product of the human imagination,
a model, derived from observations, but transcending them, with both
logical and psychological functions. It has “conceptual economy,”
replacing a great number of observations and giving them order and
coherence. In this sense, the conceptual scheme may be held as true or
19 CR, 36–41.
78 n. m. swerdlow
only as a useful model, as the two-sphere universe is still used in spher-

ical astronomy. If, however, it is held to be true, it also provides a
world-view, which, in the case of the two-sphere universe, has psycho-
logical and theological consequences, and it “explains” appearances so
that the astronomer can “understand” them, which is both logical and
psychological.
A scientist’s willingness to use a conceptual scheme in explanations is
an index of his commitment to the scheme, a token of his belief that
his model is the only valid one. Such commitment or belief is always
rash, because economy and cosmological satisfaction cannot guaran-
tee truth, whatever “truth” may mean. The history of science is clut-
tered with the relics of conceptual schemes that were once fervently
believed and that have since been replaced by incompatible theories.
There is no way of proving that a conceptual scheme is final.20
Note that Kuhn is not saying that conceptual schemes are fictitious, al-
though they may be, just that science may be right or wrong and a
good deal of it turns out to be wrong. In fact, commitment to concep-
tual schemes is essential because they are comprehensive, not limited to
what is already known, and are a “powerful tool for predicting and
exploring the unknown.” And this is a third aspect of conceptual
schemes, they are “fruitful,” providing guides for research, frameworks
for the organization of knowledge, which in turn may require exten-
sion or modification of the original conceptual schemes. What Kuhn
proposed here is quite close, both in language and in meaning, to the
function of the conceptual scheme in Conant’s definition of “the dy-
namic quality of science,” and this meaning is a central principle in
The Copernican Revolution.21
The two-sphere universe meets the criteria of being economical,
fruitful, and providing a coherent world-view, an explanation, an
understanding of the world. “The Copernican universe is itself a prod-
uct of a series of investigations that the two-sphere universe made pos-
sible. . . . The two-sphere universe is the parent of the Copernican; no
conceptual scheme is born from nothing.”22 After explaining some
alternative cosmologies in antiquity, Kuhn remarks:
Without the aid of telescopes or of elaborate mathematical arguments
that have no apparent relation to astronomy, no effective evidence for
a moving planetary earth can be produced. The observations available
to the naked eye fit the two-sphere universe very well (remember the
20 CR, 39.
21 Cf.above, 69–70.
22 CR, 41.
universe of the practical navigator and the surveyor), and there is no

more natural explanation of them. It is not hard to realize why the
ancients believed in the two-sphere universe. The problem is to dis-
cover why the conception was given up.23
The problem is in the planets, the subject of chapter 3. As an exten-
sion of the conceptual scheme of the two-sphere universe, that is, by
analogy with the two spheres of the stars and of the earth, the planets,
with motions independent of the fixed stars, were placed on a series of
spheres, also concentric to the earth, in the order of their periods.
These are in fact the “orbs” or spheres of Copernicus’s “On the Revo-
lutions of the Heavenly Spheres,” which shows that they had a very
long life, and the concentric spheres of the stars, planets, and central
earth is about as much astronomy and cosmology as laymen ever came
to know, or needed to know, through the Middle Ages. But they do not
solve the problem of the planets because the apparent motions of the
planets are so irregular. Planetary theory is likewise developed first
phenomenologically, then through the conceptual schemes devised to
account for the phenomena. The first attempt was the homocentric
spheres of Eudoxus, which were entirely qualitative but remained of
considerable importance because they were described by Aristotle and
at least implicitly became incorporated in Aristotelian cosmology (al-
though this may not have been Aristotle’s intention). On the other hand,
the theory amenable to quantitative treatment, of eccentric and epi-
cyclic circles or spheres, developed most completely by Ptolemy, never
fit that comfortably into Aristotelian cosmology. Kuhn describes both
theories, and his diagrams for Ptolemaic theory, for epicyclic, eccentric,
and equant motion, have often reappeared in other books, sometimes
with acknowledgment, sometimes without. He points out that Ptolemy’s
Almagest (ca. 150) was “the first systematic mathematical treatise to
give a complete, detailed, and quantitative account of all the celestial
motions,” and further, that it was the details of Ptolemy’s planetary
theory, such as violations of uniform, circular motion, that motivated
Copernicus’s own work. “The initial battle between Copernicus and
the astronomers of antiquity was fought over technical minutiae like
those sketched in this section.”24
However, there is an error in the exposition at this point that has
an important role in the interpretation of Copernicus’s motivation, and
that is a supposed increase in the complexity of Ptolemy’s theory through
later attempts to correct its errors or improve its accuracy by the addition
23 CR, 44.
24 CR, 73.
80 n. m. swerdlow
of epicycles and such. I do not know its original source. It is quite com-
mon in the literature, but it simply did not happen, and it no longer
appears in The Structure of Scientific Revolutions. If we leave aside the
Marāgha astronomers, to be discussed below, whose motives were pri-
marily physical, there were no structural alterations in Ptolemy’s models
and only two significant alterations of parameters, which I shall briefly
describe since they seem not to be well known but should be. The first
is a reduction of about 20 percent in the eccentricity of the solar
model, found in both Indian and Arabic astronomy, which affects only
the sun. The second is more important and affects everything. Because
Ptolemy’s observed date of equinox was a day late, there was a system-
atic error of about 21° in all longitudes at his own time and a cumula-
tive error in mean longitudes of about 2269 250 per century. By the time
of al-Battānı̄, the systematic error was 24° 209, by the time of the Al-
fonsine Tables (late thirteenth century) it was 25° 509 and by the be-
ginning of the sixteenth century it would have reached nearly 27° had
anyone still used Ptolemy’s tables directly, which, it is fair to say, no
one did. And this error in mean longitude has a variable effect in dif-
ferent parts of each planet’s synodic cycle, the period between conjunc-
tions with the sun, reaching its maximum, far greater than the mean, in
the retrograde arc. This cumulative error was periodically corrected by
finding a new, more accurate epoch, a position at a given time, for the
mean longitude of the sun, presumably by observation, and then cor-
recting the epochs and mean motions of the planets accordingly, more
likely by computation than observation. This is what one calls “reset-
ting the clock and adjusting its rate,” and other than that, Copernicus
inherited Ptolemy’s planetary theory much as he had left it. There are
of course other problems in Ptolemy’s theory, both structural and nu-
merical, but they were not of concern to anyone before Tycho and
Kepler.
With adjustment of mean motion and epoch, rather than structural
alterations to models, understood as the only significant change, we
may consider Kuhn’s following examination of how one conceptual
scheme replaces another, what may be called his “first” theory of a sci-
entific revolution.25 There are any number of conceptual schemes ca-
pable of accounting for observations already made, but not necessarily
for all possible observations. For this reason, “a scientist must believe
in his system before he will trust it as a guide to fruitful investigations of
the unknown.” But this comes at a high price, for a single observation
incompatible with his theory shows that it has been incorrect all along,
25 CR, 75–76.
and “his conceptual scheme must be abandoned and replaced.” I have

compressed, but I do not think I have misrepresented, what Kuhn calls,
in outline, “the logical structure of a scientific revolution.”
A conceptual scheme, believed because it is economical, fruitful, and
cosmologically satisfying, finally leads to results that are incompatible
with observation; belief must then be surrendered and a new theory
adopted; after this the process starts again. It is a useful outline,
because the incompatibility of theory and observation is the ultimate
source of every revolution in the sciences. But historically the process
of revolution is never, and could not possibly be, so simple as the log-
ical outline indicates. As we have already begun to discover, observa-
tion is never absolutely incompatible with a conceptual scheme.26
Note that like Conant’s definition of the dynamic quality of science, it
is the ability of a conceptual scheme to account for what is not yet
known that tests it, and Kuhn here points out the consequences of fail-
ure and of how one conceptual scheme may succeed another much as
Conant had. However, he also points out that it is not so simple. On
the one hand, as Galileo once wrote, “no greater truth either can or
should be sought in a supposition than that it corresponds to all the
particular appearances.” But on the other, a supposition, a conceptual
scheme, may be modified so that it does correspond and need not be
abandoned, so one still does not know whether it is right or wrong.
The question now, as Kuhn puts it, is “What is it that transforms an
apparently temporary discrepancy into inescapable conflict? . . . Why
do scientists hold to theories despite discrepancies, and, having held to
them, why do they give them up?” The answers to these questions, for
the specific case of the system of the world, occupy the next two chap-
ters, on why the conceptual scheme of Aristotelian cosmology was
retained, indeed, became so established that it could survive any fail-
ings that were detected, and on the first stages of what led to its replace-
ment. Let us put it another way. Copernicus’s principal innovation
could have been proposed the day after the Almagest was published,
and could have been proposed by Ptolemy himself. (Indeed, it had been
proposed more than four hundred years earlier by Aristarchus, although
apparently never taken seriously.) But it wasn’t and he didn’t. Why did
it take so long and why did it happen when it did?
These are important chapters in Kuhn’s argument, particularly
because it is in them that the intellectual history component of his
exposition is shown to be essential, that is, the intellectual factors, both
for maintaining and for doubting Aristotelian cosmology, are of greater
26 CR, 76.
82 n. m. swerdlow
importance than the scientific, or at least than the astronomical, which

hardly change at all. And it is here that he describes, and describes pre-
cisely, the Aristotelian physical world, governed, indeed formed, by
Aristotelian natural motion of the elements, which really do account
for the world as it is actually seen. There is a great deal of interest in
these chapters, and Kuhn also learned a lot about the Middle Ages,
from al-Farghānı̄’s planetary distances to Dante’s physical and spiritual
cosmology, but we shall treat them briefly. Put very simply, Aristotelian
cosmology, meaning above all the two-sphere universe, was so imbedded
in the fabric of Aristotelian natural philosophy, that it could not be
abandoned without bringing down the entire structure. For
. . . astronomy and terrestrial physics are not independent sciences.
Observations and theories developed for one become intimately en-
tangled with those drawn from another. Therefore, although difficulty
in solving the problem of the planets might have provided an astrono-
mer with a motive for experimenting in astronomy with the concep-
tion of a moving earth, he could not do so without upsetting the
accepted basis of terrestrial physics in the process. The very notion of
a moving earth would be unlikely to occur to him, because, for rea-
sons drawn from his nonastronomical knowledge, the conception
seemed so implausible.27
In addition, there is the psychological and theological belief in what
Kuhn calls “the majesty of the heavens” and the Aristotelian principle
that motion in the heavens produces effects in the air and on the earth
that provide a physical explanation of the astrological effects of the
heavens on the earth, including weather, called “mutations of the air”
and regularly predicted astrologically.28 And when Aristotelian cosmol-
ogy was adapted, not without resistance and objection, to Christian
theology, or perhaps the other way around, “any change in the plan of
the universe would inevitably affect the drama of Christian life and
Christian death. To move the earth was to break the continuous chain
of created being.”29 Thus, whatever hypothetical objections scholastic
critics may have raised to Aristotelian cosmology, to the arguments for
the central position and immobility of the sphere of the earth, to the
mobility and therefore finite dimension of the sphere of the fixed stars,
far more often than not, they found very good reasons for refuting
their own objections and concluding that Aristotle was right after all.
So much for why Aristotelian cosmology endured so long and even
grew stronger with age. Now, what, prior to Copernicus, created con-
27 CR, 86.
28 CR, 91–94.
29 CR, 113.
ditions for bringing it down? Kuhn proposes several possible causes,

among them doubts about Ptolemy’s Geography following the discov-
ery of the new world and concern with calendar reform’s drawing atten-
tion to inaccuracies in astronomical computation. But there are two he
considers the most important, about both of which I must confess to
having strong reservations. The first is the scholastic objections to Aris-
totle and hypothetical consideration of the motion of the earth in the
fourteenth century, the second the rise of Neoplatonism among human-
ists, who already had a strong anti-Aristotelian streak, in the fifteenth.30
Both are representative of currents in the historiography of science that
were receiving considerably more attention when The Copernican Rev-
olution was written than they do today, although that in itself is not an
objection to them. The first grew out of Pierre Duhem’s researches on
scholastic science of motion and his belief that scholastic objections to
Aristotle were instrumental in the formation of modern science. The
second is found in the learned and well-intentioned writings of Love-
joy, Burtt, and Koyré, who looked to Neoplatonism and Plato himself
as the source of the increasingly mathematical interpretation of nature
in the Renaissance and Scientific Revolution. One must admit that it
was preferable to Duhem’s paradoxical insistence that the origin of
modern science lay in the defense of the truths of the Catholic Church
from the incursions of Aristotle and Averroës. But that does not make
it right either. My own opinion, for what it is worth, is that neither of
these created conditions pertinent to Copernicus’s motivation, which
lay entirely within the domain of theoretical and physical astronomy
and was entirely technical, namely, a concern with the nonuniform
rotation of spheres required by the physical forms of Ptolemy’s plane-
tary models. And it was an investigation of planetary models prompted
by this problem that eventually led Copernicus, still entirely within the
technicalities of astronomy, to a solution to the still more important
problem of the order and distances of the planets and the unification of
the planetary system provided by transformations of Ptolemy’s models
explained by Johannes Regiomontanus, of whom we shall soon say
more, that led directly to the heliocentric theory. However, aside from
these few remarks, I will leave this subject to pass on to Copernicus’s
proposal of an innovation, a new conceptual scheme, that he did his
best to adapt to the universe of Aristotle but that ended up overthrow-
ing it completely.
In his exposition of De revolutionibus (1543) Kuhn concentrates
on the dedicatory preface and the introductory chapters of book 1. For
30 CR, 115–23, 127–33.

84 n. m. swerdlow
this there are two reasons, first that the remainder of the book is too
technical and mathematical for an introductory treatment, and second
that the principal arguments, or the most accessible arguments, for the
motion of the earth and the heliocentric theory are found in the open-
ing chapters. However, as Kuhn later points out, these arguments are
“profoundly unconvincing,” which may have had a good deal to do
with the incredulity and ridicule with which Copernican theory was
initially received.31 All of this is true, as is Kuhn’s well-known descrip-
tion of De revolutionibus as “a revolution-making rather than a revo-
lutionary text.”32 It stands within the ancient astronomical tradition of
the Almagest, which is almost entirely mathematical, yet contains
innovations that changed the direction of scientific thought in ways
unforeseen by its author. “It is at once ancient and modern, conserva-
tive and radical. . . . a scientific work which, although born within one
tradition of scientific thought, is the source of a new tradition that ulti-
mately destroys its parent.”33 There are two aspects of Kuhn’s analysis
on which we shall concentrate, Copernicus’s motivation and justifica-
tion for his innovations, which are quite well known because, with
some compression and alteration, they figure prominently in The Struc-
ture of Scientific Revolutions.
The first, motivation, is illustrated by Copernicus’s preface, in
which he criticizes the state of contemporary astronomy on grounds of
numerical accuracy, but more so on grounds of formal incoherence—
the comparison to a “monster” with disparate body parts is well
known—referring to the separate models for each planet with no
“fixed proportion of its parts,” meaning no common measure or unifi-
cation of the system, and the crucial problem that the models “appear
to violate the first principles, of uniformity of motion,” meaning that
in Ptolemy’s models the uniform motion of the center of the epicycle
takes place about a point, the equant point, not at the center of the
eccentric sphere.34 These are very powerful objections, for they amount
to saying that the planetary system is formally irrational and physically
impossible. Kuhn adds that the objections are also directed against
later modifications of Ptolemy’s theory, but as we have noted, there
were no such modifications, and this point is dropped in The Structure
of Scientific Revolutions. There he calls Copernicus’s preface “one of
the classic descriptions of a crisis state.”35
31 CR, 145.
32 CR, 135.
33 CR, 136.
34 CR, 136–44.
35 SR, 69.
“Crisis” is a strong word, and objections have been raised to its

application to Copernicus, but I believe it is warranted and would like
to explain why. Copernicus was not the first to make such a sweeping
indictment of astronomy. In 1464 Regiomontanus wrote a detailed
criticism of contemporary astronomy—it runs about four pages—taking
up in order the fixed stars, the planets, the sun, and the moon, and he
caustically points out errors, with the motion of the eighth sphere, that
is, the fixed stars, and the obliquity of the ecliptic, which affects the
declination of the sun, receiving particularly detailed treatment. In
the case of the planets, after noting that Saturn and Jupiter pose no
problems, the objections are based upon observations of a simple sort,
conjunctions of Mars and Venus with Saturn and of Mars with fixed
stars, the failure of Mercury to appear, times and durations of lunar
eclipses, all compared with computation from the Alfonsine Tables,
and upon implications of Ptolemy’s models, variations in apparent size
of Mars, Venus, and the moon, that are not seen to occur. All of Regi-
omontanus’s objections are observational, what is seen or not seen—
Copernicus’s objections are theoretical and physical—but their solu-
tion, particularly to the implied variation of apparent size, would nec-
essarily have theoretical consequences. It was Regiomontanus’s belief
that a thorough application of Ptolemy’s methods could be used to cor-
rect these errors, and he intended to carry out a program of observa-
tion, derivation of new elements, and then computation of new tables
and ephemerides. Unfortunately, he did not live to undertake his project,
although many hundreds of observations were later made by his asso-
ciate Bernhard Walther in Nuremberg, which were published in 1544
and were of use to Tycho and Kepler. In a way, it was Copernicus who
carried out Regiomontanus’s plan, although on a reduced scale and
with some unanticipated results. But Copernicus and Regiomontanus
have something more in common. Each was the only person of his age
to understand astronomy well enough to single out its faults. Each was
the only person of his age who had any intention to reform astronomy.
At least, no one else gave so much as a hint of either. We must not con-
fuse schoolmasters, astrologers, crack-brain Aristotelians restoring homo-
centric spheres, writers of textbooks, and computers of tables and
almanacs with serious scientists, thinking for themselves and carrying
out original research, as, say, Ptolemy or Kepler. Copernicus was the
only person in the first half of the sixteenth century who meets this
standard; the “community” of research astronomers numbered exactly
one. And if Copernicus believed that astronomy was in a sorry state
and required radical reform, then it was and it did, and it does not
make a bit of difference what anyone else thought. So crisis it was in
the eyes of the one person whose opinion mattered.
86 n. m. swerdlow
However, one caution is in order in evaluating the preface as an

explanation of Copernicus’s motivation, namely, that it was written in
1542, some thirty years after he had developed his new theory and
explained it in the Commentariolus (ca. 1510–14). There he makes no
reference to the incoherence of the separate models, the absence of a
common measure or unification of the planetary system, and concen-
trates entirely on the violation of uniform circular motion. This problem
he believed he had solved by what we now know to be an adaptation
to the heliocentric theory of models preserving the uniform rotation
of spheres developed by astronomers associated with the observatory
of Marāgha in Persia in the thirteenth century and by Ibn ash-Sh ātir of
Damascus in the fourteenth, although how he learned of their work ˙ is
still unclear.36 Hence, by his own earlier account his original motiva-
tion was concerned entirely with the physical problem of the non-
uniform rotation of the spheres in Ptolemy’s models, and in fact his
introduction of the heliocentric theory itself in the Commentariolus
appears unmotivated. Indeed, when he explains the order of the spheres
of the planets and later measures their radii, their mean distances from
the mean sun, the center of the earth’s sphere, in the common measure
of the radius of the earth’s sphere, he does not mention that he has pro-
vided a solution to the very problem of unification and common measure
that figures so conspicuously in the preface to De revolutionibus. This is
not to say that Copernicus did not realize what he had accomplished—
he most certainly did—but that unification of the system was not his
original motivation for taking up original research into the theory of
the planets, at least according to his original account. What he does
mention, and prominently, is how the heliocentric theory and motion
of the earth account for the retrogradations of the planets, although he
does not give that as his motivation either.
Hence, although this is an argument from silence in the Commen-
tariolus, it appears that his later indictment of earlier planetary theory
as incoherent did not occur to him until after he had developed the
heliocentric theory and saw that it did indeed give an unambiguous
account of the order and distances of the planets, which certainly was
one reason that he adopted it in the first place—it does not itself solve
the problem of uniform circular motion—along with its explanation of
36 The first, partial, discussion of the relation of Copernicus’s models to those of the
Marāgha astronomers appeared in the year of the publication of CR in the second edition of
O. Neugebauer’s The Exact Sciences in Antiquity, Providence: Brown University Press, 1957,
203–04. A more thorough examination came only in following years in papers by E. S.
Kennedy and his students and by Neugebauer. Copernicus’s adaptation of the models of the
Marāgha astronomers is described in appendix 2 below.
retrogradation. All of this he grasped at the time he wrote the Com-

mentariolus, but he does not there state it. Nevertheless, in writing the
preface to De revolutionibus some thirty years later, Copernicus had
every right to emphasize the incoherence of earlier planetary theory,
which in retrospect, once the consequences of the heliocentric theory
were understood, does appear an extremely serious failing, a “crisis”
that he had now corrected, and about this Kuhn’s account cannot be
more correct. But the preface cannot be taken as an accurate historical
account of his original motivation in undertaking his investigations.
Likewise in his preface he says that the motion of the earth was sug-
gested to him by the reading of ancients who held that opinion, but
this too appears to be an after-the-fact reason for approving his new
theory, an ancient pedigree no less, not a trivial argument in the six-
teenth century. Prefaces, usually written after a book is finished, are the
most purposeful and the least unvarnished sort of history, as anyone
who has written one knows.
But as Copernicus was the only person of his age to think pro-
foundly upon the difficulties of astronomy and to propose a solution
that occurred to no one else, it is hardly to be wondered that his con-
temporaries did not embrace his proposal with enthusiasm. And here
we reach the question of what it was that convinced Copernicus that
convinced hardly anyone else. What convinced Copernicus was not
accuracy, for he knew that the same parameters would produce the
same apparent motions in either system, nor was it economy alone,
although the Copernican theory does replace the epicycles of the outer
planets and the eccentrics of the inner planets with the single sphere of
the earth, but criteria that Kuhn calls “aesthetic,” a word that has
raised as many questions as “crisis”; he refers also to “harmonies” and
to “neatness and coherence,” and in The Structure of Scientific Revo-
lutions says that new theories are called “neater,” “more suitable,” or
“simpler.”37
To astronomers the initial choice between Copernicus’ system and
Ptolemy’s could only be a matter of taste, and matters of taste are the
most difficult of all to define or debate. Yet, as the Copernican Revolu-
tion itself indicates, matters of taste are not negligible. The ear equipped
to discern geometric harmony could detect a new neatness and coher-
ence in the sun-centered astronomy of Copernicus, and if that neat-
ness and coherence had not been recognized, there might have been
no Revolution.38
37 CR, 171–81; SR, 155.

38 CR, 172.
88 n. m. swerdlow
What were these geometric harmonies? There were quite a lot. Of

course the first that comes to mind is the explanation of retrograda-
tion, that a superior planet appears to move retrograde when the earth
passes it and an inferior planet when it passes the earth. And it is cer-
tainly “neater” and “simpler,” at least in terms of numbers, to replace
the epicycles of the superior planets, Mars, Jupiter, Saturn, and the
eccentrics of the inferior planets, Mercury, Venus, with the single
sphere of the earth, located between the superior and inferior planets,
all moving in the same direction about the sun. But this replacement
has a great number of consequences, and this seems as good a place as
any to list them since there is altogether too much literature today, ulti-
mately, I think, inspired by Duhem and his nonsense about “saving
the phenomena,” that holds that Copernicus had no good reasons to
believe his theory to be a true description of the world. He had very
good reasons and quite a lot of them. First, and the most important,
the Copernican theory replaces the “monster” of disparate models
governed by arbitrary relations with a unified system in which (1) the
sphere of the earth is the common measure of the spheres of all the
planets, giving the order and distances of the planets without ambigu-
ity or additional assumptions, and (2) the heliocentric periods of the
earth and the planets correspond to their order from the central sun,
while the geocentric periods of Mercury and Venus do not correspond
to their order from a central earth. Next, the same heliocentric periods
of the earth and planets (3) define the planets’ synodic periods—the
periods between successive conjunctions with the sun—which had pre-
viously been periods of revolution on the epicycle defined, for no appar-
ent reason except that they fit, by the periods between conjunctions
with the sun, a pure coincidence in geocentric theory explained com-
pletely in heliocentric.
Other arbitrary features of geocentric theory can be seen to follow
immediately from the conversion from an underlying heliocentric the-
ory. Most obviously, (4) why the radii of the epicycles of the superior
planets are always parallel to the direction from the earth to the mean
sun, indeed, (5) why the superior planets all move in the same direction
on their epicycles, another coincidence in the geocentric theory, because
their epicycles are all reflections of the earth’s heliocentric motion; (6)
why the centers of the epicycles of the inferior planets always lie in the
direction of the mean sun, because their heliocentric spheres are within
the sphere of the earth; (7) why superior planets reach opposition to
the sun, because they are outside the earth’s sphere, and (8) why infe-
rior planets reach only a limited elongation, because they are inside the
earth’s sphere; (9) why superior planets are retrograde on either side of
opposition, because that is when the earth passes them, and (10) why
inferior planets are retrograde on either side of inferior conjunction,

because that is when they pass the earth; for the same reasons (11) why
superior planets are at perigee of their epicycle when at opposition and
(12) why inferior planets are at perigee when at inferior conjunction;
and thus (13) why superior planets are brightest at opposition and (14)
inferior planets brightest near inferior conjunction.
I am not done, the list goes on: (15) why planets more distant from
the sphere of the earth, Jupiter, Saturn, and, of inferior planets, Mer-
cury, have smaller epicycles than planets closer to the earth, Venus and
Mars, because of the relative sizes of their spheres to the sphere of the
earth; (16) why planets with heliocentric periods differing greatly from
the earth’s, again Jupiter, Saturn, and Mercury, have shorter synodic
periods than planets with periods closer to that of the earth, again
Venus and Mars, because of the relative lengths of their heliocentric
periods to the period of the earth. And as a consequence of both these
relations, why more distant planets with periods differing greatly from
the earth’s have (17) shorter retrograde arcs but (18) longer periods of
retrogradation; and why closer planets with periods closer to the
earth’s have (19) longer retrograde arcs but (20) shorter periods of ret-
rogradation. (The relation of retrograde arc and time is not exactly
intuitive—the variation for a single planet is quite counter-intuitive—
and because of its very rapid motion, Mercury is an exception, with a
retrograde time as well as arc shorter than Venus’s.) For the same rea-
sons, the relative motions of the planets and the earth, why (21) more
distant planets have shorter periods of invisibility and (22) closer plan-
ets have longer periods of invisibility when near conjunction with the
sun. (Here for the inferior planets we are referring to superior conjunc-
tion; at inferior conjunction the period of invisibility of Venus, which
varies greatly, may be shorter than Mercury’s.)
It may appear that I have used too many numbers to distinguish
characteristics of superior and inferior planets and of distant and close
planets, and some are of course consequences of others, but since each
is either distinct behavior or arbitrary in the geocentric theory, but not
in the heliocentric, the numbering is not excessive. I did not even assign
a number to the explanation of retrogradation or the elimination of the
epicycles of the superior planets and the eccentrics of the inferior plan-
ets, and thus one could make the list longer. In any case, it is already
quite a long list of relations caused or explained by the order, distances,
and periods of the earth and planets in the heliocentric theory that are
either entirely absent or unexplained rules in geocentric theory. In sum-
mation, an underlying heliocentric theory explains all the characteris-
tics of the geocentric theory, but not the other way around. Anyone
who believes that Copernicus had no good reason for thinking that he
90 n. m. swerdlow
was right about the system of the world had better show that every-
thing in this list is inconsequential or coincidental. I would also recom-
mend reading Kuhn’s less compressed account, which has figures and
makes everything clear, and since I think the point is particularly impor-
tant, I have set out a description of the relation of heliocentric and geo-
centric theories in appendix 1 to this paper.
But there is yet more because Copernicus also saw relations that,
since Kepler, are known to be erroneous—they were the result of
Copernicus’s direct adaptation of Ptolemy’s models or parameters—
and we must include these also if we are to consider what convinced
Copernicus that his theory was correct. So we continue our numbering:
(23) The equations of center of the inferior planets are functions, not
of the motion of the planet, but of the earth with respect to the planet’s
apsidal line, as though these inequalities were caused by the motion of
the earth itself. The reason is that in Ptolemy’s models the equation
of center is a function of the annual motion of the center of the epi-
cycle, which Copernicus replaces by the annual motion of the earth. (24)
The variation of the radius of Mercury’s sphere is also a function of the
motion of the earth with respect to Mercury’s apsidal line, again as
though caused by the motion of the earth; this is due to Copernicus’s
adaptation of Ibn ash-Shātir’s model for Mercury. (25) The inclina-
tions of the orbital planes of˙ the superior planets vary as a function of
their synodic periods, reaching maximum inclination at opposition and
minimum at conjunction, as though the motion of the earth with
respect to the planet caused these variations just as it causes the equa-
tions of the anomaly in the same period. (26) The inclinations of the
orbital planes of the inferior planets vary with a semi-annual period as
a function of the motion of the earth with respect to the planet’s
apsidal line, reaching maximum when the earth is in the apsidal line
and minimum when the earth is a quadrant from the apsidal line, again
as though the variation was caused by the motion of the earth. (27) In
yet another component of the latitude of the inferior planets, the incli-
nations of the orbital planes also vary as a function of the earth’s dis-
tance from the planet’s apsidal line and the orbit itself must rotate
through the mean anomaly so that the point of maximum inclination
coincides with the planet whenever the earth is in the planet’s apsidal
line. Kepler did not approve of these oscillations of orbital planes—he
called them “like a monster,” which should now sound familiar—and
they are in fact due to errors in Ptolemy’s extreme latitudes, upon
which Copernicus based his theory. But what to Kepler and to us are
errors, to Copernicus were more likely further confirmations of his
theory, yet more “harmonies” of the motions of the planets with the
motion of the earth, a good reminder that it is possible to be right for
the wrong reason.
I have given more “harmonies,” relations explained by the helio-

centric theory, than Kuhn does, but he describes them at length and I
have confined myself to a bare list. Nevertheless, it is evident from
either account that Copernicus had very good reason for believing that
his new theory was the correct form of the universe, for that all of
these relations be true and yet the heliocentric theory be false, as Gali-
leo later put it, “truly surpasses my imagination.” And such aesthetic
arguments are not, as sometimes said, “soft” and “external” to science,
but, for one who understands them, are the essence of scientific and
mathematical reasoning, as every physicist knows or ought to know,
concerning which I can do no better than to recommend my colleague
S. Chandrasekhar’s essay “Beauty and the Quest for Beauty in Sci-
ence.”39 But the aesthetic arguments that convinced Copernicus did
not convince many others for the better part of a century, in part
because Copernicus himself could have explained them more clearly—
Kepler did a much better job in the first chapter of the Mysterium
Cosmographicum—but more so for reasons to be taken up below. The
Copernican Revolution was hardly complete; indeed, with Coperni-
cus’s work it had only begun, and Kuhn’s last two chapters are devoted
to the reception of Copernican astronomy, its assimilation, above all
through the work of Kepler and Galileo, and the transformation of the
world it brought about in the work of Descartes and Newton. And it is
to these two chapters that we now turn.
First the reception, which is better called resistance, for that is what
much of it was. Now Copernicus’s work was not without its admirers,
and he was called, quite rightly, a second Ptolemy as he had written a
new Almagest. It was recognized, or assumed, immediately by Erasmus
Reinhold that computation with Copernicus’s tables would probably
be more accurate than the Alfonsine Tables, although in the absence of
observational tests, of which so far there were none or almost none,
this was by no means certain. But whatever improvement there may
have been, even if only assumed, was not due to alterations in theory,
let alone to the motion of the earth, just to “resetting the clock” through
the establishment of new epochs of mean motions, as had been done
periodically for hundreds of years, for most of Copernicus’s other ele-
ments hardly differ from those of Ptolemy and the Alfonsine Tables.
This too must have been obvious to Reinhold, who went through the
39 Reprinted in S. Chandrasekhar, Truth and Beauty, Aesthetics and Motivations in

Science. Chicago: University of Chicago Press, 1987, 59–73. There are other pertinent essays
in this volume, in particular the Karl Schwartzschild Lecture “The Aesthetic Base of the
General Theory of Relativity,” 144–70, which is rather more technical. One may also note,
among other examples, Bohr’s explanation of the Balmer series in his 1913 papers on the
constitution of atoms and molecules.
92 n. m. swerdlow
colossal labor of deriving all the elements over again from Ptolemy’s
and Copernicus’s observations with meticulous (and superfluous) pre-
cision for preparing the Prutenic Tables (1551), which became the
basis of most later tables and ephemerides for nearly a century. (It is
worth noting that the results of hundreds of comparisons with obser-
vations of the Alfonsine and Prutenic Tables carried out by Tycho
many years later are ambiguous, so complex are the effects of errors in
both sets of tables; on the whole, the Prutenic Tables are somewhat
better but still far from accurate, although hardly anyone could know
that.) But Reinhold did not adopt the Copernican theory, which would
make no difference in the computation of tables anyway, and may him-
self have devised something like the Tychonic theory. Nevertheless, as
Kuhn points out, “if the decision between the Copernican and the tra-
ditional universe had concerned only astronomers, Copernicus’ pro-
posal would almost certainly have achieved a quiet and gradual victory.
But the decision was not exclusively, or even primarily, a matter for
astronomers, and as the debate spread from astronomical circles it
became tumultuous in the extreme.”40
It is certainly true, as Galileo remarked more than once, that the
opponents of Copernicus, the followers of Aristotle and Ptolemy, had
mostly never read, and those few who had did not understand, the
arguments for Copernican theory, but there were also two compelling
reasons for opposing it, one scientific, one not. The first is that to all
appearances the earth is not moving, and for nearly a century after
Copernicus it was very hard to convince anyone that the result of its
motion would not be catastrophic. Copernicus had his own explana-
tion, which he evidently considered a minimal alteration of Aristote-
lian physics, that the natural motion of a spherical body is to rotate by
virtue of its form, so that the rotation of the earth, along with the adja-
cent water, air, and anything in the air, being a natural motion, would
have no ill effects. Heaviness (gravitas), the descent of bodies with
weight toward the surface of the earth in straight lines, is the motion of
a body out of its natural place, the surface of the earth, where it would
have only the circular motion of the earth, and was explained by a
“natural inclination placed in the parts by the Divine Maker of all
things” to come together to form a globe. It is not a bad explanation
because it does account for things as they are, that is, if the earth
rotates, it explains why bodies do not fall behind or fly off but fall
down in straight lines, for which we can only be grateful to the Divine
Maker of all things. And far from overthrowing Aristotelian physics, at
40 CR, 188.
least to Copernicus’s mind it was an adaptation of Aristotelian physics

to the rotation of the earth. Galileo still held something like it,
although he turned it into a devastating onslaught on Aristotelian
physics, through the entire elimination of natural rectilinear motion,
that would have surprised Copernicus very much. But there is a para-
dox in Copernicus’s explanation. Just as it only explains the behavior
of bodies if the earth rotates, so it was only acceptable to those who
came to that opinion for Copernicus’s other reasons, that is, it was
entirely ineffective to those who were not Copernicans. It is rather like
the ontological proof, very convincing to those who believe in God,
not so convincing to those who do not.
But these physical difficulties, not entirely cleared up until Newton,
and the other scientific objections to Copernicus, as the horizon’s al-
ways bisecting the celestial sphere—Galileo slyly suggested that it
doesn’t—and the absence of detectable stellar parallax, both indicating
that the earth remains motionless at the center, were not the most im-
portant reasons for resistance. Nor was the most important reason
entrenched Aristotelianism itself, but that Aristotelianism was entrenched
in Christianity. One line of Scripture, the word of God himself, was a
more effective argument than a whole volume of Aristotle. Because sci-
ence and religion are now considered to be such good friends, and the
Church has, if somewhat grudgingly, apologized for its treatment of
Galileo, it is not fashionable to say it, but the foundation of the stron-
gest opposition was religion, and was always religion, from the begin-
ning when Luther called Copernicus a Narr, through the Church’s
condemnation of 1616, and on through the seventeenth century when
clerics in orders continued to prove that the earth does not move long
after Kepler and Galileo and even Newton had presented overwhelm-
ing evidence that it does. And Kuhn puts his finger on this directly in
twelve of the best pages on the subject I have read.41 Historians and
philosophers in recent years—ultimately if unknowingly under the
influence of Duhem—have become so enamored of scholastic distinc-
tions of probable and hypothetical reasoning that one can actually read
that Melanchthon and Bellarmine knew more about the correct prac-
tice of science than Copernicus and Galileo, even that the Church was
justified in its conduct because their proofs of the motion of the earth
were defective. But such ideas, however fashionable, are just ignorant,
or crazy. Appeals to Scripture and contrived arguments in defense of
Scripture are not refutations of scientific evidence, not now and not
then either, a point Galileo made forcefully. This was already stated
41 CR, 188–200, in particular 193.

94 n. m. swerdlow
definitively by Copernicus himself when he dismissed objections to the

motion of the earth based upon Scripture with the Olympian apho-
rism, “Mathematics is written for mathematicians.” The Church re-
sponded just as definitively in 1616 when the Copernican doctrine was
declared contrary to Holy Scripture, Galileo told that it cannot be de-
fended or held, and Paolo Foscarini’s book showing that it was not
contrary to but supported by Scripture absolutely prohibited and con-
demned. And in 1620, when De revolutionibus was “corrected,” the
passage containing Copernicus’s Olympian aphorism was ordered ex-
cised. The lines were drawn on both sides as clearly as one could wish,
and one side or the other had to give. We all know the outcome, al-
though the steps by which the Church adapted to Copernicus, to the
extent that it has adapted, were slow indeed, but that subject lies
beyond the scope of Kuhn’s book and this essay.
The scientific assimilation of Copernicus was largely the work of
two friends and one enemy. First the enemy, Tycho Brahe, who did not
approve of Copernican theory for reasons both scientific and scrip-
tural.42 One of his strongest arguments was that if stars, which were
believed to have apparent diameters of a minute or two, were removed
to a distance at which annual parallax was negligible, say, a minute or
two, they would be as large as or even larger than the earth’s orbit,
which is clearly absurd (although we now know there are such stars).
He had made a measurement of the parallax of Mars in 1582–83—
much larger than it is now known to be but then within reason—which
showed that it could come closer to the earth than the sun, thereby rul-
ing out the Ptolemaic theory, in which Mars is always more distant
than the sun, and his analysis of the motion of the comet of 1577 led
him to conclude that the heavens were not filled by rigid spheres carry-
ing the planets, for the comet would have to pass through the spheres.
This allowed him to propose, in about 1584, his well-known alternative,
the Tychonic theory, in which the planets move about the sun which in
turn moves about the central, unmoving earth.43 The Tychonic theory
retains most of the “harmonies” of the Copernican theory discussed
42 CR, 200–09.
43 My colleague Howard Margolis has shown, contrary to what everyone (including me)
has believed, that the Tychonic theory is compatible with solid spheres, for the sun can carry
the entire system around the earth without intersecting the sphere of Mars, which is always
around the sun, although the sun’s path does intersect the path of Mars. The belief that the
spheres intersect turns out to be an illusion, although a very natural one, in interpreting
the standard diagram of the Tychonic theory, which disappears if the circular path of the sun
is deleted from the diagram and the sun with the planetary system is seen as moving about
the earth. See Margolis’s note “Tycho’s Illusion and Human Cognition,” Nature 392:857 (30
Apr 1998).
above and eliminates all the physical and scriptural objections, al-
though, as Kuhn notes, it has its own incongruities: the planets are
badly off center, the center of the universe is not the center of most
celestial motions, it is hard to imagine a physical mechanism that can
produce the required motions, and, something Kuhn does not mention
but does show in his figure, the superior and inferior planets move
around the sun in opposite directions, a significant point, overlooked
in whole volumes on Tycho, that he must have caught in drawing the
figure. Hence, it does not meet the aesthetic criteria met by the Coper-
nican theory, as shown by Kepler, who marshaled no less than eighteen
arguments against it in his Epitome of Copernican Astronomy (1618–
21). But as Kuhn also notes, Tycho was an unwitting friend of Coper-
nicus, for “any break with the tradition worked for the Copernicans,
and the Tychonic theory, for all its traditional elements, was an impor-
tant break.” Tycho’s elimination of solid spheres is not Copernican—
Copernicus still retained them, although this point has been disputed—
but it is definitely not Aristotelian as then understood, and the location
in the heavens of the new star of 1572 and of several comets was also a
serious blow to Aristotelian cosmology. “Somehow, in the century
after Copernicus’ death, all novelties of astronomical observation and
theory, whether or not provided by Copernicans, turned themselves
into evidence for the Copernican theory. That theory, we should say,
was proving its fruitfulness.”44
It is difficult to say which of Copernicus’s two friends, Kepler and
Galileo, was the greater Copernican. But each was a Copernican in a
different way, and Kuhn catches the distinctions perfectly. One thing
common to both, however, was a commitment to Copernican theory,
in the sense that Kuhn holds that commitment to a conceptual scheme
is essential to its extension, for it to be “fruitful.”45 First Kepler, whose
contributions are of three kinds.46 First, and the best known, are what
we now call the ellipse and area laws of planetary motion, which cor-
rectly describe a planet’s first inequality, its nonuniform motion about
the sun, developed in the Astronomia nova (1609), to which we must
add the new elements of the planetary orbits, derived over a period of
many years from Tycho’s observations, and the computation of the
Rudolphine Tables (1627), in accordance with the new theory and new
elements, which put all previous astronomical tables in the shade
(although that was not immediately obvious and their use is extremely
44 CR, 208.
45 Cf.above, 78.
46 CR, 209–19, 244–47.
96 n. m. swerdlow
laborious). The models for the first inequality of planetary motion of

Ptolemy, Copernicus, and Kepler are described in appendix 2. It has
been said that Kepler’s laws do not necessitate the Copernican theory,
but anyone who believes this has not thought the matter through. For
the Ptolemaic alternative would require the planets to move on ellipti-
cal epicycles with eccentricities exactly proportioned to the eccentricity
of the sun and apsidal lines holding fixed directions, for the superior
planets parallel to the apsidal line of the sun, while the foci of the ellip-
tical epicycles in turn move in elliptical orbits about the earth, all
motions obeying the area law; and the Tychonic version would require
the apsidal lines of the ellipses of the planets to hold fixed directions as
their foci at the sun moved in an ellipse about the earth, again, every
motion following the area law. In order to keep the apsidal lines in
their proper directions, it is further required that the epicycles, or
ellipses about the sun, rotate in the direction opposite to their motion
about the earth or sun, the rotations obeying the area law no less.
Kepler dismissed this kind of thing as hopelessly implausible and cum-
bersome at an early stage in the Astronomia nova when he showed that
the earth’s heliocentric motion cannot be on a simple eccentric but
must have an equant, a good thing after all, which would have similar
baroque consequences in the other two systems. So Kepler’s laws alone
are powerful evidence for the Copernican theory since it is all but
impossible to place them in any other, and thus the Copernican theory
shows its capacity for simplification in a way that Copernicus could
never have imagined.
His second contribution was physical, and although this aspect of
Kepler’s work has only recently received the attention it deserves, Kuhn
saw its importance and wrote a good description of it.47 With the
spheres gone, an explanation of how the heavenly bodies move at all
was imperative, and as planets, both individually and as a system, are
faster as they are closer to the sun and slower as they are farther from
the sun—a relation that does not even exist with respect to the earth in
geocentric theory as individual planets, far from being fastest, are actu-
ally retrograde when closest to the earth—Kepler took this to be a
“motive virtue,” an “image” of the body of the sun, extending out-
ward from the sun, its effect decreasing with distance, and causing the
planets to move in circles as the solar body itself rotated, which he
inferred from his theory before its demonstration from the motion of
sun spots. Here is something only a committed Copernican could think
47 CR, 214–16, 244–47. The most complete treatment is B. Stephenson, Kepler’s Physical
Astronomy, New York: Springer, 1987 (rpr. Princeton, 1994).

of, showing the importance of commitment to a conceptual scheme if it

is to be extended. And it was the direct relation between distance from
the sun and the time required by a planet in each small arc of its motion,
the foundation of his physical theory, that led Kepler to the area law,
the sum of the times in each arc, as a means of computing the planet’s
motion, for areas are only a means of computation and the physics
remains the relation of time in each arc to distance. His theory is more
complicated than this, as the effect of the virtue to move the planet
must decrease linearly with distance from the sun and a magnetic-like
attraction and repulsion between the sun and planet is required to
account for the orbit’s taking an eccentric elliptical rather than a con-
centric circular form, but the point to be made here is that Kepler’s
physical theory is unthinkable in anything but the heliocentric theory
and thus, whether it be ultimately right or wrong, is yet another exten-
sion of Copernican theory.
Kepler’s third contribution may have been to him the most impor-
tant of all, and that was his truly vast extension of Copernicus’s aes-
thetic arguments for the heliocentric theory through the discovery of
countless relations undreamt of by Copernicus. With one exception
these have turned out to be false or not significant, but that does not
mean that they were not significant to Kepler and to demonstrating the
Copernican theory. Kuhn calls them Neoplatonic; my own preference
is to say that Kepler, whose favorite name for them was “archetypes,”
was discovering God’s plan of the universe. The first, in the Mysterium
cosmographicum (1596), is that the intervals between the distances of
the six planets from the sun found by Copernicus correspond to the
inscription of the five regular solids in a series of spheres, which also
shows why there are six planets. The correspondence is close, although
not exact, and offers no explanation of why the orbits should be eccen-
tric; but in the Harmonice mundi (1619) Kepler showed that the angu-
lar velocities of the planets at aphelion and perihelion produced by
those eccentricities correspond closely to ratios of harmonic, that is,
musical, intervals. These are “harmonies” indeed—although not all are
consonant—and he found many more between the motions of adjacent
planets and in the entire planetary system. And since the ratios of har-
monic intervals are those of the divisions of a circle by the sides of
constructible regular polygons, God evidently planned the universe
according to the immutable laws of geometry. Finally, he discovered
what is known as his third law, that the squares of the periods of the
planets are proportional to the cubes of their mean distances from the
sun, that the periods are as the three-halves power of the distances,
which he even found a way of explaining by his physical theory and
archetypes of the weights and sizes of the planets. Needless to say, none
98 n. m. swerdlow
of these relations makes sense, most do not even exist, in anything but
the heliocentric theory, and Kepler considered them overwhelming evi-
dence for the truth of Copernican theory and for his discovery of God’s
plan. Again, it is possible to be right for the wrong reason, but right or
wrong, Kepler’s is the boldest cosmology ever conceived, and only a
committed Copernican could have conceived it.
For anyone who studied Kepler’s work with an open mind and
could understand it, which is not easy, Kepler had proved the Coperni-
can theory. He had, as Kuhn says, solved the problem of the planets.
But here now is a paradox. Galileo, who had no confidence, perhaps
no interest, in Kepler’s reasoning and evidence, and did not have the
patience to study Kepler either, also set about to prove the Copernican
theory in his own way, which has nothing to do with Kepler’s way, for
Galileo was only interested in doing things his own way. That is the
way it is with original genius. It is all so well known, but it is worthy of
its fame. The telescopic observations of the surface of the moon
showed that the moon is a solid body with a rough surface like the
earth, the explanation of the secondary light of the moon, the “ashen”
light of the dark regions in the crescent phases, showed that the earth is
a body shining by reflected light from the sun, in turn reflecting light to
the moon as the moon reflects light to the earth. Hence the earth down
here and the moon up there are alike, and the moon certainly moves.
The stars were numerous beyond all counting, and shown to be so
small by the telescope that they could be removed to distances at which
parallactic effects would be undetectable without being any larger than
the sun. So much for Tycho’s argument. The satellites of Jupiter, the
first additions to the planetary system discovered since the most remote
antiquity, showed that the moon could move around the earth as the
earth moved around the sun, as Jupiter was certainly moving about
something. The phases of Venus showed that it unquestionably moved
about the sun, which could be extended by induction to Mercury and
then to the superior planets since, aside from reaching opposition to
the sun, their motions do not differ from that of Venus. If all the plan-
ets, why not the earth? And the highly irregular spots that appeared
and disappeared on the surface of the sun showed, even more than the
occasional comet or nova, that the heavens were far from unchanging
and, as Kuhn remarks, “worst of all, the motion of the spots across the
sun’s disc indicated that the sun rotated continually on its axis and thus
provided a visible paradigm for the axial rotation of the earth.” 48
48 CR, 221–22.
Do Galileo’s observations prove the Copernican theory? Not quite,

because they can be explained in other ways, although perhaps not
plausibly, particularly in the case of the relation of the earth and the
moon, and because they do not prove the motion of the earth, al-
though they certainly show it to be possible and remove some impor-
tant objections. “Therefore,” Kuhn observes, “the telescope did not
prove the validity of Copernicus’ conceptual scheme. But it did provide
an immensely effective weapon for the battle. It was not proof, but it
was propaganda.”49 But Galileo knew the Copernican theory was cor-
rect, he was determined to prove it, and to prove it in his own way, re-
lying on nothing more than Copernicus, and little enough of that, and
his own ingenuity. That is the subject, the purpose, of the Dialogue on
the Two Great Systems of the World, Ptolemaic and Copernican (1632),
the single greatest sustained argument in the entire literature of science.
Galileo was not supposed to do this, he had been specifically forbidden
to do this, but he thought he could get away with it when everyone, or
at least everyone who mattered, saw that his demolition of Aristotle
and proof of Copernicus were unanswerable. And knowing exactly for
whom he was writing, he explained the essential points, such as the
Copernican account of the order of the planets, retrograde motion, and
even the seasons, practically in words of one syllable. Kuhn treats the
Dialogue briefly, and we therefore will not say much about it here.
There is only one point I wish to make. Galileo really did believe he
had proved the two motions of the earth, from the apparent motion
of sunspots, as a result of the sun’s rotation and the annual motion of
the earth, and from the tides, which he believed could occur “within
the limits of nature” in no other way than by a daily variation in the
speed of water in seas caused by the diurnal and annual motions of
the earth and then modified by the local conditions of different bodies
of water. The latter does not stand up to Newtonian mechanics, and
perhaps even to Galilean mechanics, but that simply shows that Galileo
too, like Copernicus and Kepler, could be right for the wrong reasons.
“The triumph of Copernicanism was a gradual process, and its rate
varied greatly with social status, professional affiliation, and religious
belief. But for all its difficulties and vagaries it was an inevitable pro-
cess. At least it was as inevitable as any process known to the historian
of ideas.”50 With Kepler and Galileo the astronomical part of the
Copernican theory was essentially proved, even if the proof was not
49 CR, 224. I would give more weight to Galileo’s evidence, which appears to me all but
conclusive.
50 CR, 227.
100 n. m. swerdlow
yet absolutely complete, and alternatives reduced to near impossibility

if not absurdity. Of course there were still legions of Simplicios, of phi-
losophers, who would not read Kepler and Galileo, and may not have
been able to understand them if they tried—Galileo truly believed that
the principal reason for opposition to Copernicus was ignorance, sheer
ignorance born of stupidity, and Galileo was usually right—but it is
safe to say that anyone who understood the issues and still opposed
Copernicus was constrained to do so for theological reasons. There
were certainly non-astronomical arguments against Copernicus that
had not been answered. But what this really means is that if the non-
astronomical sciences, as physics and theology, contradicted Coperni-
cus, if they could be used as arguments against Copernicus, the fault
belonged not to Copernicus, but to physics and theology, a point that
still seems to be inadequately appreciated. This is the subject of Kuhn’s
last chapter, the adjustment in other sciences necessitated by the victory
of Copernicanism. “Until those adjustments were made, the Coper-
nican Revolution was incomplete.” How do such adjustments take
place? Innovations are initially introduced to solve special problems in
a single scientific specialty in spite of their conflict with common sense,
physical intuition, and the basic concepts of other sciences. But with
continued use in one area, the new conception gains a larger scientific
function and ceases to be an ad hoc device for describing the known,
and becomes a basic tool for explaining and exploring nature, at which
point it cannot be restricted to a single scientific specialty as nature
ought not display incompatible properties in different fields. “Every
fundamental innovation in a scientific specialty inevitably transforms
neighboring sciences and, more slowly, the worlds of the philosopher
and the educated layman.”51
Once accepted, Copernicanism gave rise to a new set of problems,
as the dimensions of the much larger universe it necessitated, the
source of the motion of the planets, and the unification of celestial and
terrestrial physics that it seemed to imply if the earth is a planet. In
each case, the early stages differ little from the Aristotelian world; thus,
for Copernicus, the two-sphere universe still holds except that the
sphere of the stars is much larger and the earth an insignificant distance
from the sun at the center, the planets are still carried by spheres, and
Aristotelian physics is altered only slightly to make possible a moving
earth. But the “sphere” of the stars really has no necessary function in the
Copernican universe, especially once all the spheres are gone, and Kuhn
sees Copernicanism as giving a scientific basis to what had previously
51 CR, 229–30.
been nothing more than occasional speculation in scholasticism, Neo-

platonism, and atomism about an infinite universe scattered with stars,
which is surely correct because speculation of this kind increased dra-
matically near the end of the sixteenth century among early Coperni-
cans. “Copernicanism therefore allowed a new freedom to cosmological
thought, and the result was a new speculative conception of the uni-
verse that would have horrified both Copernicus and Kepler.” 52 Indeed
it would, for it was strange speculation, as shown by Bruno’s infinite
heavens filled with stars and spiritually populated planets, and also by
Thomas Digges’s well-known figure of the Copernican system (1576),
which he calls “the most auncient doctrine of the Pythagoreans,” of
nested spheres no less, surrounded by stars extending infinitely upward,
with its remarkable, and seldom read, caption, almost in verse:
This orbe of starres fixed infinitely up extendeth hit self in altitude
sphericallye, and therefore immovable, the pallace of foelicitye, gar-
nished with perpetuall shininge glorious lightes innumerable, farr
excellinge our sonne both in quantitye and qualitye, the very court of
coelestiall angelles, devoyd of greefe and replenished with perfite end-
less ioye, the very habitacle for the elect.
What this is supposed to mean I can only guess, as I guessed about
where to put commas, but the innumerable perpetual shining glorious
lights inhabited by celestial angels and the elect are a long way from
the Newtonian universe and from what Galileo saw with his telescope.
One thing, however, is certain, that they are the fruitfulness of Coper-
nicanism. The principal study of this entire subject, Alexander Koyré’s
From the Closed World to the Infinite Universe, appeared in the same
year as The Copernican Revolution; much of its story makes Bruno
and Digges seem models of sobriety. Curiously, all this speculation
about an infinite universe belongs less to Copernican cosmology, or to
astronomical science of any kind, than to adapting Christian theology
to Copernicanism, that is, Copernicanism is taken as correct and it is
the obligation of Christian theology to adapt, just as Dante adapted
Christian theology to the Aristotelian heavens in the Paradiso. Dante,
of course, was writing an allegory and he knew it, but I do not think
we can attribute such caution to Bruno, Digges, and their followers.
Yet we can certainly understand their difficulties, for if it is taken seri-
ously, Copernicanism still poses problems to theology. Indeed, one may
wonder whether theology has ever taken in the significance of the
removal of the earth from the center of the universe, let alone all that
has followed from it.
52 CR, 232.
102 n. m. swerdlow
Kuhn’s last chapter is puzzling, at least, it puzzled me at first. It

would seem that from Kepler and Galileo one could go in a straight
line to Newton, who, in the Principia (1687), from nothing more than
Kepler’s laws, his own laws of motion, and an extension of the kind of
mechanics used by Galileo in the Two New Sciences, aided by far more
advanced mathematics, showed that the entire system of planets and
satellites is governed by a single inverse-square centripetal force, identi-
cal to the force of gravity, thereby solving the problem of the planets
completely; and that the center of gravity of the system, about which
the sun and planets move, is very near the sun, thus proving the helio-
centric theory once and for all beyond the least shadow of a doubt.
Not that Newton needed to consult either Kepler or Galileo directly,
although there is evidence that he did; the very little he required was
readily available elsewhere and, like Galileo, he did everything his own
way. But that straight line may be just logic, and logic is not necessarily
history. Instead, Kuhn detours into the apparently remote subject of
corpuscularism, for which there turns out to be a very good reason, as
we shall see.53 One of the outstanding problems created by Coperni-
canism and by the disappearance of spheres was what causes the plan-
ets to move, and to move in nearly circular paths. Galileo still held, like
Copernicus but without spheres, that circular motion is natural and
continues on its own, and Kepler had his own explanation, his own
physics, using immaterial forces that act over distances to do what
needs to be done. But neither was possible or even rational to, above
all Descartes, who sought to account for these motions through com-
pletely general principles governing the motions of all bodies beginning
with the smallest corpuscle. The motion of the corpuscle, if nothing
affected it, could only be straight, and, to make a moderately long
story short, by assuming a great number of corpuscles, filling all of
space and pushing each other, and an even greater amount of hand
waving, he concluded that they would arrange themselves into vast cir-
culations, vortices, that would nicely carry the planets around the sun.
They would also account for, among much else, comets, gravity, tides,
and the transmission of light, unifying the physics of the heavens and
the earth from entire planets to single corpuscles under the same prin-
ciples. So arose one of the most attractive, influential, and wrong-
headed theories, indeed conceptual schemes, in the entire history of
science.
53 In 1951–52 Kuhn had published four papers in Isis on atomism in seventeenth-century
chemistry, and in “A Discussion with Thomas S. Kuhn,” 170–72, he discusses his early
interest in atomism.
Kuhn reviews Descartes’s theory, more carefully and seriously than

I have, and also describes theories that employ forces, Kepler’s, with
one force to move the planet in a circle and an attractive and repulsive
force to bring it within and return it to the circle, Giovanni Borelli’s, in
which one force moves the planet and a second, attractive force regu-
lates its distance from the sun, and Robert Hooke’s, in which a single
force directed to the sun deflects the planet from motion in a straight
line into some kind of closed curve.54 Hooke’s model in particular is
one of applied mechanics, like a pendulum or projectile, the sort of
thing Galileo would think of, showing that here too the division between
celestial and terrestrial physics has been crossed. And Hooke went even
farther, supposing that all bodies in the planetary system attract each
other with the same force that causes bodies to fall on the earth, that
this force causes bodies to depart from motion in a straight line to a
curve, as a circle or an ellipse, and that the strength of the force varies
with distance from the center of the attracting body. “Now what these
several degrees are I have not yet experimentally verified; but it is a
notion, which if fully prosecuted as it ought to be will mightily assist
the astronomer to reduce all the celestial motions to a certain rule,
which I doubt will never be done true without it.”55
Speculation, however, was one thing, and demonstration another,
and the latter was beyond Hooke. But demonstration is exactly what
Newton, who probably knew all of this before it ever occurred to
Hooke, did, in a rough way perhaps as early as 1666 when he discov-
ered from Kepler’s third law that the centripetal force acting on the
planets is inversely as the square of the distance from the sun and,
assuming that the same force acts between the earth and the moon, is
none other than the force of gravity that causes heavy bodies to
descend at the surface of the earth. The original demonstrations may
have been rough, but he later made them more rigorous and also dem-
onstrated that any centripetal force whatever produces the uniform
description of areas by the line joining a moving body to the center of
force of Kepler’s second law, and that the elliptical orbit of Kepler’s
first law implies an inverse-square centripetal force at the focus of the
ellipse. “These mathematical derivations were without precedent in the
history of science. They transcend all the other achievements that stem
from the new perspective introduced by Copernicanism.” 56 One may
add, as noted before, that Newton’s demonstration, from computing
the masses of the planets with satellites compared to the mass of the
54 CR, 237–52.
55 CR, 254, quoting Hooke.
56 CR, 256.
104 n. m. swerdlow
sun, that the center of gravity of the planetary system, about which the
sun and planets move, must be very near the sun proved the Coperni-
can theory beyond any doubt. Newton’s demonstrations in the Prin-
cipia are summarized in appendix 3. Very simply, if Newton’s laws of
motion are true and the periods and distances of planets and satellites
are those given by observation, the Copernican theory is true. It is as
certain as that.
However, two difficulties remained, and they were both incompati-
bilities with corpuscularism. By Newton’s demonstration, that the cen-
tripetal force acting on the moon is identical to the force of gravity, the
attraction of the earth, whether on the moon or a stone, is inversely as
the square of the distance from its center. But the attraction of the earth
must, by the principles of corpuscularism, be the sum of the attractions
of all the corpuscles that make up the earth. And while distance from
the center may be a reasonable approximation to this for the remote
moon, either because its distance from all parts of the earth is nearly
the same or because its distance from the center may be taken as an
average, it is far from clear that anything like this applies to a nearby
stone, for which the attractive forces of parts of the earth at different
distances must differ very greatly. Only in 1685, when working on the
Principia did Newton succeed in proving that the inverse-square rule
holds rigorously for all distances from the center of a sphere beyond its
surface, that is, a “corpuscle” placed anywhere without the spherical
surface of the earth and attracted inversely as the square of the distance
from every corpuscle of the earth is attracted inversely as the square of
the distance from the center, as though every corpuscle making up the
earth were at its center.
That surprising discovery, which at last rooted gravity in the individ-
ual corpuscles, was the prelude and perhaps the prerequisite to the
publication of the Principia. At last it could be shown that both
Kepler’s Law and the motion of a projectile could be explained as the
result of an innate attraction between the fundamental corpuscles of
which the world machine was constructed.57
And that is the reason for Kuhn’s detour through corpuscularism
instead of going straight from Kepler to Newton, and he could not
have been more right, more profoundly right, than to do it the way he
did. For it is this series of demonstrations, which became Principia
l.71–76 and their corollaries, and which Newton himself, who hardly
ever commends anything, called “worthy of note” (Quod est notatu
dignum), that unifies the mechanics of the heavens and the earth under
57 CR, 258; the word surprising cannot be too strongly emphasized.

a single set of laws. And only after these demonstrations, in corollaries

to 1.76, does Newton first state what we call the law of gravity in
something close to its familiar form, that, for motive attractions,
weights, of spherical bodies with spherically symmetrical distributions
of matter, upon each other: the attractions are directly as the product
of the spheres, meaning their quantities of matter, their masses, and
inversely as the square of the distances between their centers. Note the
restriction, rigorously necessary but usually overlooked in stating
the law of gravity, to spherical bodies with spherically symmetrical
distributions of matter, of which there may be any number. Newton,
naturally, did not miss it.58 Kuhn saw the significance of Newton’s
demonstrations clearly, and his last chapter was built around it.
At last the crumbling Aristotelian universe was replaced by a compre-
hensive and coherent world-view, and a new chapter in man’s devel-
oping conception of nature was begun. . . . The construction of
Newton’s corpuscular world machine completes the conceptual revo-
lution that Copernicus had initiated a century and a half earlier.
Within this new universe the questions raised by Copernicus’ astro-
nomical innovation were at last resolved, and Copernican astronomy
became for the first time physically and cosmologically plausible.59
One last problem remained, reducing gravity itself to corpuscular-
ism, or to mechanical causes of any kind, which Newton himself was
neither the first nor the last to attempt without success. Perhaps, we
may say, that is where the Copernican Revolution left off and a new
revolution, an entirely new way of thinking, was required, a revolution
that began only in our own century and is not yet completed.
58 In the preceding section 11, on the motion of bodies attracting each other by centripetal
forces, Newton considers proportionality to mass, but in bodies without size. The closest he
comes to the modern formulation is in 3.7–8, where the restriction is still to spheres with
spherically symmetrical distributions of matter. The case of direct linear force is considered
in 1.77–78 and the difficult extensions to arbitrary forces and non-spherical figures in 1.79–
93; all are present in the first edition of the Principia, which is likewise worthy of note.
59 CR, 260–61. This is the beginning of Kuhn’s conclusion, “The New Fabric of
Thought,” CR, 261–65. Rather than condense or comment on it, for it must be read, I would
simply recommend reading it.
106 n. m. swerdlow
Appendix 1
Relation of Ptolemaic, Copernican,
and Tychonic Theories
Geocentric and heliocentric theories produce the same apparent motions

of planets, and are related to each other by simple geometric transfor-
mations. These are different for superior planets, Mars, Jupiter, Saturn,
which reach opposition to the sun, and inferior planets, Mercury, Venus,
which reach a limited elongation, and thus we consider them separately.
In the following descriptions, we assume the simple case of circular orbits
in the plane of the ecliptic, the circle described by the annual motion of
the sun, with the earth or sun at the center, which is all that is required.
Figure 1a shows the Ptolemaic model for a superior planet, in
which the earth is O, the center of the epicycle C, and the planet P; the
sun S9 lies in the direction OS9, but its distance is not specified. Center
C moves about O in a circle of radius OC 5 R, completing a revolu-
tion in the planet’s zodiacal period, return to the same longitude in the
ecliptic. Planet P moves about C on an epicycle of radius CP 5 r, com-
pleting a revolution in the planet’s synodic period, return to conjunction
Figure 1. Ptolemaic, Tychonic, and Copernican Theories of the Superior Planets

with the sun, such that the planet is at apogee A at conjunction, mov-
ing in the same direction as C, at perigee B at opposition, moving in
the direction opposite to C, and in general that the radius of the epi-
cycle CP is always parallel to the direction OS9 from the earth to the
sun. The ratio r/R is derived from observation, as are the zodiacal and
synodic periods, and the two radii r and R are unique to each planet,
with no common measure of distance in the system of planets. Conse-
quently, the distances and even the order of the planets are undeter-
mined without further assumptions, as that order corresponds to period.
It follows from the conditions just described that the apparent motion
of the planet is fastest near conjunction, when the motions of P and C
are in the same direction, and slowest, in fact retrograde, near opposi-
tion, when the apparent motion of P in the direction opposite to C
exceeds the motion of C. Likewise, the planet is farthest from the earth
and faintest near conjunction and closest to the earth and brightest
near opposition. Now, what is the reason for all these relations of the
motion of the planet to the motion of the sun? Because that is the way
the model is designed; it conforms to these relations, all of which are
correct, but does not explain them. In order to explain them, we must
transform the model from geocentric to heliocentric, which shows why
these relations are necessary.
Complete parallelogram OCPS and let the sun be at S, which lies in
the direction OS9; the apparent direction of the planet OP is the diago-
nal of the parallelogram. There are two significant transformations, the
Tychonic shown in Figure 1b and the Copernican in Figure 1c, in both
of which we show parallelogram OCPS, from which they are formed,
and its diagonal, the apparent direction OP. We shall explain the
Copernican theory first as it is the more familiar. The earth on a circle
of radius SO 5 r and the planet on a circle of radius SP 5 R move
around the sun in the same direction through their respective zodiacal
periods, one year for the earth, longer for a superior planet. The syn-
odic period, return of the planet to conjunction with the sun, is now
defined by the return of the earth to A on line PS extended, and oppo-
sition takes place when the earth is at B. It is now obvious why the
planet is fastest, farthest from the earth, and faintest near conjunction,
slowest, in fact retrograde, closest to the earth, and brightest near
opposition. Indeed, retrogradation occurs near opposition because the
faster earth is then passing the slower planet. It is also apparent why in
the geocentric theory the radii of the epicycles of all three superior
planets are parallel to the direction from the earth to the sun, because
the motion of each planet on the epicycle is in fact the single motion of
the earth about the sun. Were the motion on the epicycle measured
from a fixed direction in the ecliptic rather than the moving direction
108 n. m. swerdlow
of A, the period of the planet’s motion would be one year, the same as
the earth’s motion about the sun measured from a fixed direction.
What were only rules in the geocentric theory, perfectly correct rules,
are seen to have causes in the heliocentric theory in the relative motions
of the earth and planet. The ratio r/R, stated as R/r, is the same as before,
but now the radius of each planet’s orbit R is known in the common
measure of the radius of the earth’s orbit r, which means that the order
and distances of the planets may be determined securely by observation.
The Tychonic theory is shown in Figure 1b, in which the sun S
moves around the earth O in a circle of radius OS 5 r with its zodiacal
period of one year, and the planet P moves about the sun S in a circle of
radius SP 5 R, completing a revolution in its synodic period, return to
conjunction with the sun at apogee A, in the direction opposite to the
motion of the sun about the earth; opposition is reached at perigee B.
It is these opposite motions that make the Tychonic theory less intu-
itively obvious than the Copernican. Note that the sun carries the orbit
of the planet about the earth like an epicycle, which in fact it is, a large
epicycle. Since the motion of P is opposite to the motion of S, the zodi-
acal motion of the planet, which is their difference, is slower than the
motion of the sun and the period of the planet is longer than one year.
Were the motion of the planet measured from a fixed direction in the
ecliptic rather than the moving direction A, it would complete a revo-
lution about the sun in its zodiacal period, moving in the same direc-
tion as the sun. Still, since the orbit of the planet is carried by the sun,
the synodic motion in the opposite direction is its real heliocentric
motion. Just as in the other theories, the planet is fastest, farthest from
the earth, and faintest near conjunction at A, slowest, in fact retro-
grade, closest, and brightest near opposition at B. Retrogradation
occurs near opposition when the apparent motion of the planet, oppo-
site to the motion of the sun, exceeds the motion of the sun carrying
the orbit of the planet. Admittedly, this is less obvious, harder to visu-
alize, in the Tychonic theory than in either the Ptolemaic or Coperni-
can. Nevertheless, as in the Copernican theory, what were only rules in
the Ptolemaic theory, relations of the motion of the planet to the sun,
have causes in the Tychonic theory in the relative motions of the sun
and planet. And likewise the ratio R/r gives the radius of the planet’s
orbit R in the common measure of the distance between the sun and
earth, the radius of the sun’s orbit r, so again the order and distances of
the planets may be determined securely by observation.
The relation of the theories of the inferior planets is more straight-
forward. The essential property of the motion of the inferior planets is
that they reach only a limited elongation from the sun. In the Ptolemaic
model, shown in Figure 2a, the center of the epicycle S, moving on a
Figure 2. Ptolemaic, Tychonic, and Copernican Theories of the Inferior Planets
circle of radius OS 5 R, always lies in the direction OS9 from the earth
to the sun, the distance of which is not specified, completing a revolu-
tion around the earth in one year. The planet P moves on an epicycle of
radius SP 5 r, completing a revolution in its synodic period, return to
superior conjunction at apogee A, where P moves in the same direction
as S, and reaching inferior conjunction at perigee B, where P moves in
the direction opposite to S. Again the ratio r/R is derived from observa-
tion, and the two radii r and R are unique to each planet, with no com-
mon measure in the system of planets, so that order and distances are
undetermined; Ptolemy’s order was Mercury, Venus, sun, but other
orders were also used, even locating Venus above the sun. The appar-
ent motion of the planet is fastest near superior conjunction, when the
motion of P and S are in the same direction, and slowest, in fact retro-
grade, near inferior conjunction, when the apparent motion of P in the
direction opposite to S exceeds the motion of S. The planet is farthest
from the earth near superior conjunction and closest near inferior con-
junction, but the brightness does not quite correspond to distance because
the inferior planets show phases like the moon, as was first discovered
110 n. m. swerdlow
for Venus by Galileo with the telescope. Now, why should it be that the
centers of the epicycles of the inferior planets always lie in the direction
of the sun? Because that is the way the model is designed in order to
produce limited elongation from the sun, but to explain this essential
property, we must transform the model from geocentric to heliocentric.
In the Tychonic form in Figure 2b, we simply assume that the sun is
located at S, the center of the epicycle, which had already been sug-
gested in antiquity, so OS 5 R is the radius of the sun’s orbit around
the earth and SP 5 r the radius of the planet’s orbit around the sun. All
the relations of the motion of the planet to the sun in the Ptolemaic
model are preserved, but now it is obvious why the center of the epi-
cycle always lies in the direction of the sun and why the planet reaches
only a limited elongation from the sun, because the sun is at the center
of the epicycle, which is smaller than the orbit of the sun about the
earth, and why the planet moves retrograde near inferior conjunction,
because then the apparent motion of the planet, opposite to the motion
of the sun, exceeds the motion of the sun carrying the orbit of the
planet. Further, r/R shows the ratio of the radius of the planet’s orbit r
to the sun’s orbit R, so the order and distances may be determined by
observation in the common measure of the radius of the sun’s orbit,
just as for the superior planets.
In the Copernican form in Figure 2c, we let the earth O move
around the sun in an orbit of radius SO 5 R, completing a revolution
in one year, and the planet move around the sun in an orbit of radius
SP 5 r, completing a revolution in its own heliocentric period, which is
less than one year, both motions measured from a fixed direction. The
synodic period, the difference of the heliocentric motions of the planet
and the earth, is the same as in the Ptolemaic and Tychonic models, the
planet reaching apogee A at superior conjunction and perigee B at infe-
rior conjunction. All relations of the Ptolemaic model with regard to
speed and distance are preserved. It is evident why the center of the
epicycle lies in the direction of the sun and why the planet reaches only
limited elongation from the sun, because the sun is at the center of the
planet’s orbit, which lies within the earth’s orbit, and further, why
the planet moves retrograde near inferior conjunction, because then the
faster planet is passing the slower earth. Again the ratio r/R gives
the ratio of the planet’s orbit r to the earth’s orbit R and so determines
the order and distances in the common measure of the radius of the
earth’s orbit, as for the superior planets.
These then are the transformations of the models for the superior
and inferior planets in the three theories, in which it can be seen that
the Ptolemaic follow rules, all of which are correct, while the Coperni-
can and Tychonic give the causes of these rules. As noted earlier, an
underlying heliocentric theory explains all the characteristics of the

geocentric theory, but not the other way around. Of the relations
described earlier, the “geometric harmonies,” numbers 1 through 14
follow from the transformations in these figures alone, and 15 through
22 also require the empirically derived ratios of distances and periods.
It is sufficient in summation to consider the defining properties of the
motion of superior and inferior planets:
(1) The radii of the epicycles of the superior planets are always par-
allel to the direction from the earth to the sun because the motion of
the planet on the epicycle is nothing other than the motion of the earth
about the sun or of the sun about the earth.
(2) The centers of the epicycles of the inferior planets always lie in
the direction of the sun because the planet moves about the sun in an
orbit smaller than the orbit of the earth about the sun or of the sun
about the earth.
There is, however, nothing in these relations to exclude either
the Copernican or Tychonic theories. It may seem peculiar that in the
Tychonic theory the synodic motions of superior and inferior planets,
take place in opposite directions with respect to the direction from the
earth to the sun, but when measured from a fixed direction in the eclip-
tic all the motions are in the same direction, just as in the Coperni-
can theory. Kepler has no less than eighteen arguments in support of
Copernican and against Tychonic theory, some very good, some very
strange. Galileo’s demonstrations of the necessity of the annual and
diurnal motions of the earth through the motions of the tides and of
sunspots would be decisive against Tycho, which he surely had in
mind, if the demonstrations were sound. The explanation of the tides,
as it turned out, is not, but the explanation of sunspots is, if not deci-
sive, which it just might be, far superior to any other. Still, in the course
of the seventeenth century, the Tychonic theory continued to lose
ground to the Copernican, and by the time Newton refuted it defini-
tively by showing that the common center of gravity of the planetary
system, about which everything including the earth moves, is very near
the sun, almost no one was left to defend it. Newton’s demonstration is
shown in appendix 3.
Appendix 2
Relation of the Models for the First Inequality
of Ptolemy, Copernicus, and Kepler
The heliocentric theory is concerned with what is called the second

or solar inequality of the planets, the nonuniformity in their apparent
112 n. m. swerdlow
motions due to the motion of the earth about the sun, which is the sub-
ject of appendix 1. The first or zodiacal inequality is the planet’s own
nonuniform motion about the sun. It is a more difficult problem than
the second inequality and its correct description is given by Kepler’s
first two laws:
1. The planet moves in an ellipse with the sun at a focus.
2. The line joining the planet to the sun describes an area propor-
tional to time.
These are shown in Figure 3a in which we let the apsidal line be AB
with center M and place the sun S at a focus distant from M by the
eccentricity e. The planet P moves on the ellipse such that the line SP
describes an area ASP proportional to time, that increases uniformly
with time. Note that in the figure the eccentricity and ellipticity are
greatly exaggerated compared with the orbits of the planets, which
are nearly circles with small eccentricities. Now, the line EP joining the
second, empty focus of the ellipse to the planet describes an angle AEP
Figure 3a, 3b. Models of the First Inequality of Kepler and Ptolemy
that happens to be very nearly proportional to time, that is, the angular
motion of P about E is very nearly uniform. This is the principle under-
lying Ptolemy’s model. For in Figure 3b we superimpose Kepler’s
model of an ellipse and Ptolemy’s model of a circle, now shown helio-
centrically so there is no epicycle, and we see that the circle is the major
auxiliary circle of the ellipse, the circle constructed on the major axis.
In Ptolemy’s model the planet P9 moves on the circle with center M uni-
formly with respect to point E, later called the “equant” point, that is,
angle AEP9 increases uniformly with time. The purpose of the model is
to separate the center of constant distance M from the center of uni-
form motion E such that the variation of distance SP9 is determined by
the single eccentricity SM 5 e and the inequality of direction, angle
SP9E, by the double eccentricity SE 5 2e. But point E is none other
than the empty focus of the ellipse, about which, as noted, the motion
of the planet is very nearly uniform. The difference between the two
models, the difference in the directions SP and SP9, is the small angle
PSP9, which, in the case of Mars with a rather large eccentricity of
about 0.1, reaches a maximum of about 41– e 2 < 89 of arc. Ptolemy’s
model is about as close as one can come to the effect of Kepler’s laws,
in both direction and distance, without knowing them, and Kepler
remarked that the difference of 89, which it took him years of hard
work to eliminate, led the way to reforming all of astronomy.
Kepler believed the physical cause of planetary motion to be a force
from the sun, which rotates with the body of the sun and moves the
planet such that the time required by the planet to describe each small
arc of its orbit is proportional to its distance from the sun; the areas
proportional to time were only a method of computation, of summing
the times in each small arc proportional to the distances. Before finding
the ellipse, Kepler had applied this method to motion in a circle, as is
also shown in Figure 3b, where the planet is on the circle at P̄ and area
ASP̄ is proportional to time; this also produces an error, angle PSP̄, of
about 89 compared to motion in the ellipse, but in the opposite direc-
tion. That difference of 89 also gave Kepler a lot of trouble to eliminate.
The relation between motion of the planet P̄ on the circle and P on the
ellipse is straightforward; draw an ordinate P̄N to the apsidal line, and
the planet P is at the intersection of the ordinate and the ellipse.
The motivation of Kepler’s model for the first inequality was both
accuracy and physical, a mechanics of celestial motion necessitated by
Tycho’s elimination of solid spheres carrying the planets, a conclusion
Kepler seems to have reached on his own. Copernicus’s concern with
the first inequality was also physical, but the opposite, a model com-
patible with the uniform rotation of solid spheres, which was violated
by Ptolemy’s separation of the center of distance and the center of
114 n. m. swerdlow
uniform motion such that, in Figure 3b, a sphere with center M is

required to rotate uniformly about point E not at its center. There is no
way a rigid sphere can do this without being displaced from center M.
As we have noted, Copernicus was not the first to be concerned with
this physical problem in Ptolemy’s model as astronomers associated
with the observatory of Marāgha in Persia in the thirteenth century
had also considered it, and among the solutions they devised some are
identical to those used by Copernicus. Although the path of transmis-
sion of their work to Europe is not certain—the most likely is through
Italy in the fifteenth century—Copernicus’s models for the planets,
aside from the heliocentric form, and for the moon are so close that he
must have known of them in some way. So many details of Coperni-
cus’s models are identical to those of his Arabic predecessors that inde-
pendent invention is next to impossible. The model described below is
identical to that of Mū’ayyad ad-Dı̄n al-’Ur dı̄ (d. 1266) and Qu tb ad-
˙
Dı̄n ash-Shı̄rāzı̄ (d. 1311); in the earlier Commentariolus Copernicus˙ used
a form with two epicycles identical to that of Ibn ash-Sh ātir (d. 1375).
Copernicus’s and Ptolemy’s models for the first inequality ˙ are
superimposed heliocentrically in Figure 3c, in which we let the mean
sun, the center of the earth’s sphere, be at S and the apsidal line be SA
with A the highest apsis. In Ptolemy’s model the planet P moves on a
circle of radius R with center M, uniformly about the equant point E
through angle k, the mean eccentric anomaly, also shown in Figure 3b.
Figure 3c. Models of the First Inequality of Ptolemy and Copernicus

The direction from the mean sun to the planet is SP. In Copernicus’s
model, the center of a small epicycle with center C moves on a circle of
radius R with center M9, uniformly about M9 through k, and the planet
P9 moves on the epicycle, uniformly about C through k, such that
when C is in the apsidal line the planet lies at the lowest point of the
epicycle on the line M9C closest to M9. The direction from the mean
sun to the planet is SP9. Provided that SM9 5 32– e and CP9 5 12–e, P9
always lies on line EP just beyond P and coincides with P in the apsidal
line. Thus P9 also moves uniformly about the equant point E, which is
the purpose of Copernicus’s model, maintaining uniform motion about
the equant point of Ptolemy’s model but now by means of two uniform
circular motions, of C about M9 and of P9 about C. Hence, Copernicus
does not, as is commonly said, eliminate the equant, but retains it
strictly. However, the path of the planet is no longer a circle about M,
but a figure that has what Kepler calls an “exorbitation” from a circle,
that lies just outside the circle and coincides with it only in the apsidal
line. Nevertheless, Copernicus’s model is nearly identical to Ptolemy’s
in the direction of the planet, which is what can be observed. For
Mars, with an eccentricity of about 0.1, the greatest difference in the
directions SP and SP9 is about 39, near the limit of accuracy of Tycho’s
observations but so masked by other errors, as in parameters, as to be
undetectable.
Unlike the heliocentric theory and the motion of the earth, Coper-
nicus’s model for the first inequality was generally approved because of
its strict adherence to the uniform circular motion of spheres. Tycho,
even without solid spheres, continued to use it in the double-epicycle
form, used by Copernicus in the Commentariolus, also shown in the
figure. Here an epicycle with center C9 and radius C9C 5 32– e moves uni-
formly about S through k, C moves about C9 in the opposite direction
through k such that C9C is always parallel to the apsidal line, and CP9
moves through 2k measured from C9C. The location of P9 is identical
in the two models. But Kepler, who did approve of the heliocentric the-
ory, did not approve of Copernicus’s model in any form, and remarked
in the Astronomia nova that Ptolemy’s model is absurd if solid spheres
are assumed but quite suitable and probable physically if they are denied,
and Copernicus’s model is absurd if solid spheres are denied but barely
tolerable physically if they are assumed, right on target as usual. In
developing his own planetary theory, he began by rejecting Coperni-
cus’s model and returning to Ptolemy’s because the violation of uniform
motion, such that the planet moves slower when farther from the sun
and faster when closer, was the very foundation of his own physical
theory of planetary motion. So in the case of the first inequality, Ptolemy
was right after all.
116 n. m. swerdlow
Appendix 3
Newton’s Demonstration of the Force
of Gravity and the Definitive Proof
of the Heliocentric Theory
It is no exaggeration to call these the most important demonstrations

in the history of science. They are distributed between the mechanics of
centripetal forces in book 1 of the Principia and their application to the
System of the World in book 3. We shall consider them together as
briefly as we can without losing their content. Even though a good
many propositions are involved, Newton’s proofs are direct and eco-
nomical. The evidence for the laws of motion is empirical although
qualitative. To give a modern quantitative example of nearly inertial
motion, each second the earth travels about 30 km in its orbit about
the sun and departs from a straight line by less than 3 mm, which can be
entirely accounted for by the effect of the mass of the sun, which in turn
is known independently by its effect on every other planet. This is very
close to inertial motion and is strong evidence for the first two laws.
In 1.1 Newton shows from the first two laws of motion and ele-
mentary geometry of similar triangles, both of finite and indefinitely
small width, that the radius joining any body compelled to move in an
orbit to an unmoving center of force describes areas proportional to
the times. In 1.2 he shows the inverse, that any body that moves in a
curve and by a radius drawn to a point describes areas proportional to
the times is urged by a centripetal force tending to that point. The
proofs are completely general and apply to any central forces whatever.
As we earlier noted, Kepler believed the true physical cause of plane-
tary motion to be a force from the sun that moves the planet such that
the times required by the planet to describe each small arc of its orbit
are proportional to its distances from the sun. The uniform description
of areas, the proportionality of areas to times, was for Kepler only a
method of computation, of summing the times in each small arc pro-
portional to the distances. Newton showed instead that the planet
moves by its inertia in a straight line, that the departure from motion
in a straight line is due to a central force, and that the description of
areas proportional to times is not just a method of computation but
fundamental, following directly from the first two laws of motion. It is
with good reason that the uniform descriptions of areas are the first
propositions in the Principia. These propositions are applied in book
3. In 3.Phenomena 5, he points out that it is well known to astrono-
mers that the planets by radii drawn to the earth in no way describe
areas proportional to times, as they move both direct and retrograde,
but with respect to the sun they move nearly uniformly and by radii
drawn to the sun do describe areas proportional to times. In 3.2 he

draws the conclusion that the forces by which the planets are drawn
away from rectilinear motion and retained in their orbits are directed
to the sun. This itself is very strong evidence for the heliocentric theory.
For the next series of demonstrations, we shall use some elemen-
tary mathematics. In 1.4 and cor. 1–2, Newton finds the centripetal
force of bodies moving in a circle, that is, their acceleration toward the
center of the circle. We let f be force, which, by the second law of
motion, is proportional to acceleration a, v be orbital velocity, p
period, and r distance from the center of the circle. Then, by consider-
ing the perpendicular from the tangent of a small arc to the circle, that
is, the departure from rectilinear motion, the centripetal force or accel-
eration, f , a 5 v2/r 5 (2pr/p)2/r , r/p2. Hence, by 1.4 cor. 6, if
according to Kepler’s third law, p , r3/2 or p2 , r3, so f , a , r/p2 ,
r/r3 5 1/r2, the force is inversely as the square of the distance. This is
obviously of some importance. In 3.Phenomena 1, 2, 4, Newton shows
that the periods of the satellites of Jupiter and Saturn and of the pri-
mary planets about the sun are as the three-halves power of their dis-
tances, that p , r3/2. Hence, by 3.1–2 the forces by which the satellites
of Jupiter and Saturn and the primary planets about the sun are drawn
off from rectilinear motion and retained in their orbits are directed to
the centers of Jupiter, Saturn, and the sun and are inversely as the
squares of the distances from those centers.
There are two further demonstrations that the forces keeping plan-
ets in orbit are inversely as the square of the distance. In 1.11 it is
shown that if a body revolve in an ellipse, the centripetal force toward
a focus of the ellipse is inversely as the square of the distance, and this
is extended in 1.12–13 to motion with respect to the focus of a hyper-
bola and parabola. Since the planets are known to move in ellipses, this
is further evidence for an inverse-square centripetal force. Further, the
planets describe ellipses about the sun, not about the earth, so the
inverse-square force is directed to the sun, not to the earth. Then in
1.45 it is shown that where the exponent of the distance r of a centrip-
etal force is less than 22 the apsidal line of an orbit moves forward, in
the direction of the body’s motion, where the exponent is greater than
22 the apsidal line moves backward, opposite to the direction of the
body’s motion, and where the exponent is exactly 22, that is, as 1/r2,
the apsidal line does not move and the orbit is stable. Now the apsidal
lines of the planets hardly move at all with respect to the fixed stars, and in
3.2 Newton concludes that the aphelia of the planets at rest show the cen-
tripetal force directed to the sun to be as 1/r2 very exactly, for the slight-
est deviation from 1/r2 would produce a notable motion of the apsides
in a single revolution and a very great motion in many revolutions.
118 n. m. swerdlow
This can be seen in the case of the moon, for the motion of the
apsidal line in one revolution is about 3°, from which it follows, by 1.45
cor. 1, that the exponent of the centripetal force is (360/363)2 2 3 5
22243 4
––– . This is very close to 22, and the slight discrepancy can be
accounted for, not by a difference in the centripetal force, but by per-
turbation by the sun, although in the Principia Newton was not able to
account completely for the motion of the moon’s apsidal line in this
way. Nevertheless (3.3), this is very good evidence that the centripetal
force retaining the moon in its orbit about the earth is directed to the
earth and is in fact as 1/r2. In 3.4 Newton shows that this force is iden-
tical to the force of gravity that causes bodies to descend at the surface
of the earth. We shall give the demonstration briefly and in modern
units; Newton uses the distance the moon departs from a straight line,
one-half the acceleration, in one minute, but we shall use the entire
acceleration in one second. From the period and distance of the moon,
about 27.32 days and 60 radii of the earth or 384,000 km, it is found
that v < 1 km/sec and the acceleration of the moon toward the earth
am 5 v2/r < 0.272 cm/sec2. Hence, since the centripetal force acting on
the moon is as 1/r2, the acceleration at the surface of the earth ae 5
r2am 5 602 am < 9.8 m/sec2, precisely the experimentally verified accel-
eration of a falling body at the surface of the earth. The centripetal
force acting on the moon is therefore none other than the force of grav-
ity, since otherwise bodies falling to earth would be acted upon by both
forces and would fall twice as fast. And thus, the force of gravity must
also vary inversely as the square of the distance and, as shown in 3.5–
7, is itself the inverse-square centripetal force proportional to mass act-
ing throughout the system of planets and satellites.
There remains the definitive proof of the heliocentric theory. Empiri-
cal evidence is given in 3.Phenomena 3 that the planets surround the
sun with their orbs, namely, the phases of Mercury and Venus, Mars’s
appearing gibbous at quadratures, and Jupiter and Saturn always full,
but these appearances do not prove that the earth moves around the
sun. This requires a physical proof, which is also definitive for all
the planets. By 1.57, an application of the third law of motion, two
bodies that attract each other describe similar figures about their com-
mon center of gravity and about each other, and their distances from
their common center of gravity are inversely as their masses. In the case
of the sun and planets, where is the common center of gravity? The
masses of two bodies can be compared if their accelerative effects on
other bodies at known distances can be compared. Now the gravita-
tional force or acceleration is directly as the mass and inversely as the
square of the distance, that is, f , a , m/r2. But, as we have seen, the
centripetal force or acceleration f , a , r/p2, and it is now known that
these are the same force. Hence, the mass m , fr2 , (r/p2)r2 5 r3/p2.
Therefore, if we are given the period and distance of a planet from the
sun, pp and rp, and the period and distance of a satellite from a planet
ps and rs, we have for the mass of the sun mS , rp3 /pp2 and for the mass
of the planet mp , r 3s /p2s . The ratio of the masses is therefore
m r3s ⁄ p2s rs 3 p 2
-------p = ------------------- =  ----  -----p- .
mS r 3p ⁄ p2p  rp   p s 
In 3.8 cor. 1–2, using the periods and distances of satellites of Jupiter,
Saturn, and the earth compared to the period and distance of Venus,
Newton finds the ratios of the masses of the planets with satellites to
the mass of the sun: Jupiter 1/1067, Saturn 1/3021, earth 1/169,282. 60
In 3.12 he concludes that the common center of gravity of Jupiter and
the sun falls in a point a little above the surface of the sun, the common
center of gravity of Saturn and the sun falls in a point a little below
the surface of the sun, that even when all the planets are aligned on the
same side of the sun, the common center of gravity of all is scarcely one
solar diameter from the center of the sun, and that in other configura-
tions it is always less. With modern figures, Newton’s conclusions are
still correct. Hence, the common center of gravity of the planetary sys-
tem, about which every planet and the sun itself move, is very near the
sun and nowhere near the earth. And with respect to the sun, by 1.65
and 3.13, the planets move very nearly in ellipses with the sun at a
focus and describe areas very nearly proportional to the times, the
qualification “very nearly” due to perturbations by other planets. I in-
clude this summary of Newton’s proof since even today one can read
that the heliocentric theory has not been proved, that it is only a theory.
Obviously, this is nonsense, as is the common idea that the detection of
stellar parallax was necessary to prove the heliocentric motion of the
earth. I note also in conclusion that anyone who believes the law of
gravity and the heliocentric theory to be true without knowing New-
ton’s demonstrations or something equally rigorous does not, if I may
paraphrase Conant, “understand” science, but accepts cosmology as
part of a revealed religion.
60 These numbers, from the third edition, differ in the earlier editions and in Newton’s
own annotated copies, and although Jupiter and Saturn have only small errors there is a
notable miscomputation for the earth, concerning which see R. Garisto, “An Error in Isaac
Newton’s Determination of Planetary Properties,” American Journal of Physics 59 (1991),
42–48.
120 n. m. swerdlow
References
Writings of Thomas S. Kuhn cited in this essay:
The Copernican Revolution. Planetary Astronomy in the Development of Western
Thought. Cambridge, Mass.: Harvard University Press, 1957.
The Essential Tension, Selected Studies in Scientific Tradition and Change. Chicago:
University of Chicago Press, 1977.
The Structure of Scientific Revolutions. Third Edition. Chicago: University of Chicago
Press, 1996.
Writings of James B. Conant pertinent to the historical method of teaching natural science
at Harvard:
On Understanding Science. An Historical Approach. New Haven: Yale University
Press, 1947.
The Growth of the Experimental Sciences. An Experiment in General Education.
Progress Report on the Use of the Case Method in Teaching the Principles of the
Tactics and Strategy of Science. Cambridge, Mass.: Harvard University Press,
1949.
Science and Common Sense. New Haven: Yale University Press, 1951.
With L. K. Nash, ed. Harvard Case Histories in Experimental Science. 2 vols. Cam-
bridge, Mass.: Harvard University Press, 1957.
My Several Lives. Memoirs of a Social Inventor. New York: Harper & Row, 1970.
Other works consulted or pertinent to this essay:

“A Discussion with Thomas S. Kuhn.” A discussion between Thomas S. Kuhn and
Aristides Baltas, Kostas Gavroglu, Vasso Kindi. Neusis 6 (1997), 143–98;
reprinted in T. S. Kuhn, The Road since Structure, ed. J. Conant and J. Hauge-
land. Chicago: University of Chicago Press, 2000, 255–323.
General Education in a Free Society. Report of the Harvard Committee. Introduction
by J. B. Conant. Cambridge, Mass.: Harvard University Press, 1945.
Cohen, I. B. “A Harvard Education.” Isis 75 (1984), 13–21.
Freedman, M., ed. Roosevelt and Frankfurter, Their Correspondence, 1928–1945.
Boston: Little, Brown, 1967.
Greene, J. D. The Tercentenary of Harvard College. A Chronicle of the Tercentenary
Year, 1935–1936. Cambridge, Mass.: Harvard University Press, 1937.
Hershberg, J. G. James B. Conant. Harvard to Hiroshima and the Making of the
Nuclear Age. New York: Knopf, 1993.
Hufbauer, K. “Thomas Kuhn’s Discovery of History (1940–1958),” presented at the
conference on The Legacy of Thomas S. Kuhn at the Dibner Institute, November
1997. To appear in Historical Studies in the Physical and Biological Sciences.
Keller, P. Getting at the Core. Curricular Reform at Harvard. Cambridge, Mass.: Har-
vard University Press, 1982.
Kuhn, T. S. “Professionalization Recollected in Tranquillity.” Isis 75 (1984), 29–32.
Shirley, J. W. “The Harvard Case Histories in Experimental Science: The Evolution of
an Idea.” American Journal of Physics 19 (1951), 419–23.
Westman, R. S. “Two Cultures or One? A Second Look at Kuhn’s The Copernican
Revolution.” Isis 85 (1994), 79–115.

An Essay On Thomas Kuhn's First Scientific Revolution

Uploaded by

Copyright:

Available Formats

An Essay On Thomas Kuhn's First Scientific Revolution

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Essay On Thomas Kuhn's First Scientific Revolution

Uploaded by

Copyright:

Available Formats

An Essay on Thomas Kuhn’s

First Scientific Revolution,

T HE C OPERNICAN R EVOLU TION , Planetary Astronomy in the

honor, the honorary chairman of the United States Harvard Tercentenary

splendidly. He later wrote to Felix Frankfurter, “I told the boys afterwards

1 On the Harvard Tercentenary, I have followed Conant, My Several Lives, 146–56,

Redbook remains a pertinent and thoughtful evaluation of education in America, at once

Conant described the new method of instruction in three Terry Lec-

for actually engaging in scientific research. Although Conant’s purpose

idea is therefore not only its success in correlating the then-known

on the elasticity of the air, concentrating on Boyle, the second, very

technology, the Case Histories are confined to pure science, perhaps as

Copernican Revolution, which he continued to do each year before he

of science. Nevertheless, it is the work of an extraordinarily creative

only as a useful model, as the two-sphere universe is still used in spher-

universe of the practical navigator and the surveyor), and there is no

and “his conceptual scheme must be abandoned and replaced.” I have

importance than the scientific, or at least than the astronomical, which

ditions for bringing it down? Kuhn proposes several possible causes,

30 CR, 115–23, 127–33.

“Crisis” is a strong word, and objections have been raised to its

However, one caution is in order in evaluating the preface as an

retrogradation. All of this he grasped at the time he wrote the Com-

37 CR, 171–81; SR, 155.

What were these geometric harmonies? There were quite a lot. Of

inferior planets are retrograde on either side of inferior conjunction,

I have given more “harmonies,” relations explained by the helio-

39 Reprinted in S. Chandrasekhar, Truth and Beauty, Aesthetics and Motivations in

least to Copernicus’s mind it was an adaptation of Aristotelian physics

41 CR, 188–200, in particular 193.

definitively by Copernicus himself when he dismissed objections to the

laborious). The models for the first inequality of planetary motion of

Astronomy, New York: Springer, 1987 (rpr. Princeton, 1994).

of, showing the importance of commitment to a conceptual scheme if it

Do Galileo’s observations prove the Copernican theory? Not quite,

yet absolutely complete, and alternatives reduced to near impossibility

been nothing more than occasional speculation in scholasticism, Neo-

Kuhn’s last chapter is puzzling, at least, it puzzled me at first. It

53 In 1951–52 Kuhn had published four papers in Isis on atomism in seventeenth-century

Kuhn reviews Descartes’s theory, more carefully and seriously than

57 CR, 258; the word surprising cannot be too strongly emphasized.

a single set of laws. And only after these demonstrations, in corollaries

Geocentric and heliocentric theories produce the same apparent motions

Figure 1. Ptolemaic, Tychonic, and Copernican Theories of the Superior Planets

Figure 2. Ptolemaic, Tychonic, and Copernican Theories of the Inferior Planets

underlying heliocentric theory explains all the characteristics of the

The heliocentric theory is concerned with what is called the second

uniform motion such that, in Figure 3b, a sphere with center M is

Figure 3c. Models of the First Inequality of Ptolemy and Copernicus

It is no exaggeration to call these the most important demonstrations

drawn to the sun do describe areas proportional to times. In 3.2 he

Other works consulted or pertinent to this essay:

You might also like