Over the years I have been asked many times if I would recommend Arthur Koestler’s The Sleepwalkers: A History of Man’s Changing Vision of the Universe (1959, ppb 1964) and if I wouldn’t, what would I recommend instead? My answer to the first question is a clear, yes/no, maybe, don’t know! The answer to the second one is complex.
In order to explain my answers, I need to fill in a bit of background. The Sleepwalkers is a single volume (six hundred plus pages in the paperback version) popular history of western cosmology from six hundred BCE through to Isaac Newton. I acquired and read my copy, which I still have, in the late 1960s and it had a massive impact on my development as an ongoing historian of science. In fact, I would, without hesitation, list it as one of the most important books in my life that helped shape my intellectual development.
Having described it as a popular history of western cosmology, I will have to qualify that somewhat. It started out originally as a biography of Johannes Kepler, and in the final book the section on Kepler, The Watershed, covers almost two hundred of those six hundred pages. The book opens with sixty pages devoted to the ancient Greeks from Pythagoras to Ptolemy. A further thirty to the European Middle Ages before we arrive at Copernicus, who gets about one hundred pages of his own, under the title The Timid Canon. The section on Kepler includes his interactions with Galileo and the final section is ninety pages that cover The Burdon of Proof, the Trial of Galileo, and The Newtonian Synthesis. The book closes with a thirty-page epilogue, followed by a quite extensive selected bibliography and fifty pages of endnotes. So, what we really have is not a popular history of western cosmology but rather an extensive biography of Johannes Kepler within an extended historical context.
When I started writing this blog post, I intended to talk about Koestler’s book from memory but taking it in my hands to write the paragraph above I decided spontaneously to reread it. This has proved to be a very interesting experience. I first read it about fifty years ago, and as I said above it had a major impact on me, I can’t remember if I have reread part or all of it in the intervening five decades. The first thing that hit me on rereading is just how much I have learnt about the topic over those fifty years. The second is just how bad some sections of Koestler’s book are viewed from the standpoint of the knowledge that I have acquired since first reading it. However, in Koestler’s defence, the book was researched and written in the second half of the 1950s and published in 1959 and was largely based on the then current expert opinions. Modern history of science was still in its infant phase at the time. What follows in a review of the book from my current perspective based on our actual knowledge of the histories of cosmology and astronomy.
Reading it today, I find Koestler’s take on astronomy/cosmology in Ancient Greece quite frankly bizarre. He starts with Pythagoras to whom he attributes the discovery that the world is a sphere and then makes a big deal out of the Pythagorean number theory, claiming it is the application of mathematics to the study of nature and that the Pythagorean’s were the inventors of science. Way over the top in my opinion. He then claims Heraclides of Pontus as a Pythagorean and attributes the Capellan geo-heliocentric system, in which Mercury and Venus orbit the Sun, which in turn orbits the Earth, to him. To be fair this was a common attribution at the time which has been since then debunked by Bruce Eastwood. It was apparently based on a misinterpretation of Macrobius. Martianus Capella is the earliest known propagator of this system. Heraclides did, however, propagate a geocentric model with diurnal rotation. Koestler now jumps to Aristarchus of Samos, whom he also views as a Pythagorean, who famously propagated a heliocentric system, the original account of which we don’t have, and which apparently found almost no adherents. Having built a ‘perfect’ Pythagorean cosmology, Koestler now lays into Plato, Aristotle and co, ridiculing the axiomatic uniform circular motion of the planets, claiming it to be one of the greatest mistakes in the history of science. In this he is not actually wrong. He rubbishes the Eudoxan geometrical model only granting that it was better than the deferent/epicycle model of Ptolemy. He dismisses Ptolemy as a waste of space. Here endeth Koestler’s gospel on Ancient Greek cosmology. His take can be roughly summarised as you had the truth served up on a plate from the Pythagoreans–Pythagoras, Heraclides, Aristarchus–and you threw it away causing two thousand years of groping around in the dark.
It comes as no surprise that Koestler’s take on the Middle Ages is even worse that his take on Antiquity. To start off the Middle Ages are very definitely the “Dark Ages”! Admittedly a fairly standard view in the 1950s. It goes almost without saying that Sleepwalker’s Koestler is a flat earther, that is he states very clearly that people in medieval Europe believed that the world was flat. He introduces the works of Lactantius and Cosmas Indicopleustes as proof of this and claims that medieval Europe didn’t return to a belief in a spherical Earth until about 1000 CE. Actually, in one sentence he says the ninth century, in another around 900 CE and in a third about 1000 CE. He acknowledges that Bede in the seventh/eighth century knew that the world was a sphere but seems to think he had little influence. He has a major rant about Neoplatonism, which he thinks dominated European thought until the twelfth century and was totally anti-scientific. He even quotes the circular medieval mappa mundi as proof that the world was thought to be flat, whereas they were round in order to signify that the world was a sphere. He acknowledges that Islamic scholars translated Greek knowledge into Arabic but brings the totally false claim that they only preserved it but added absolutely nothing of value to it before returning it to Europe in the High Middle Ages. He also emphasises that the Islamic translators and preservers were Persians, Jews, Christians etc but not Arabs! It is interesting reading all this because it was mainstream history of science thought in the 1950s when Koestler wrote his book.
Koestler accepts the theory that Western thought can be characterised by an alternating dominance between the Platonic and Aristotelian philosophies. First came Plato then came Aristotle, then came Neoplatonism, and he sees a glimmer of hope in the Renaissance of Aristotelian thought in the High Middle Ages championed by Thomas Aquinas. Hope that is shattered by the rigid non-scientific scholasticism of the medieval university. I could go on but lets just agree that Koestler’s view of medieval science is to put it mildly dated and should be consigned to the dustbin of history. Having then signalled the end of the Middle Ages and the beginning of the scientific revolution with a new change to Plato, this time Pythagorean Neoplatonism, also a dated concept, Koestler now turns his attention to Copernicus.
Koestler does a real hatchet job on the poor Nicolaus. However, unlike his distorted and dated views of Antiquity and the Middle Ages this can be regarded and positive and refreshing. There was a vast hagiography in popular texts both concerning Copernicus and his De revolutionibus and as far as I know Koestler was the first to seriously challenge the perceived image of the man and the book that gave us the Copernican Revolution, the Astronomical Revolution, kicked off the Scientific Revolution, heralded the Modern World and, and, and…
Koestler’s title for his Copernicus section says it all, The Timid Canon. Far from being the revolutionary, reforming hero, who bravely kicked over two thousand years of stagnant geocentric cosmology, freeing the world from this error in perception to usher in a new scientific age, Copernicus was a dithering reactionary, who lacked the courage of his convictions to present his idea to the world. As Koestler shows. Copernicus was not trying to break with tradition, he was, on his own account, trying to rescue Aristotle’s dictate that the planets orbit the Earth with uniform circular motion, which he saw as breached by Ptolemy’s use of the equant point. We still don’t know how Copernicus came to make the transition from a geocentric system to a heliocentric one to rescue the uniform circular motion axiom, but we do know that he was looking back to Ancient Greece rather than forward to the future. Koestler was seriously criticised for deflating Copernicus’ popular image; you attack a cultural icon at your own risk. Today, we might not be quite so harsh as Koestler but in general the view that Copernicus was at best a reluctant revolutionary, if not a reactionary, is generally accepted.
Koestler’s source for his biography of Copernicus is Leopold Prowe’s two volume German biography Nicolaus Coppernicus (1883), which was at the time he was writing the most recent and comprehensive biography available, so it is fairly accurate. However, his biography of Rheticus is fairly inaccurate. For example, it includes the following bizarre statement:
…as a young man he studied at the Universities of Zurich, Wittenberg, Nuremberg, and Goettingen.
Rheticus went to school, not university, in Zurich, and then studied at the University of Wittenberg. The University of Goettingen was first founded in the 1730s about 160 years after Rheticus’ death and there never was a University of Nuremberg!
Koestler is also not very complimentary about De revolutionibus, he points out, quite correctly, that if Copernicus wants it to be a realistic picture of the cosmos, how does he think that all of his deferents and epicycles function in the real world? Most notorious is Koestler’s heading for his first section on De revolutionibus, The Book that Nobody Read. The historian of science, Owen Gingerich, who particularly disliked Koestler’s portrayals of Copernicus and Galileo, used the heading as the title of his popular book[1]describing his experiences during his thirty-year odyssey compiling his census of the extant copies of the first and second editions of De revolutionibus[2].
In his census, Gingerich recorded all the marginalia made by readers of the copies of the De revolutionibus that he surveyed; there were many and some were very extensive. He argued, correctly, on the basis of the evidence that those who bought De revolutionibus did in fact read it and so Koestler was wrong. There is a slight ironical twist to Gingerich’s argument. Koestler’s book is expressly a history of cosmology but one of Gingerich’s findings was that readers extensively annotated the technical, mathematical sections of the book, whilst completely ignoring the sections on cosmology.
However, Gingerich’s argument doesn’t actually address Koestler’s reasons for his, admittedly somewhat hyperbolic, heading; Koestler brings two. Firstly, he notes the comparatively small number of copies of De revolutionibus printed. He states:
Its first edition, Nuremberg 1543, numbered a thousand copies, which never sold out. It had altogether four reprints in four hundred years: Basle 1566, Amsterdam 1617, Warsaw 1854, and Torun 1873.
Then in comparison he writes about other major astronomy works:
The most popular among them was the textbook by a Yorkshireman, John Holywood (died 1256), known as Sacrobosco, which saw no less than fifty-nine editions. The Jesuit father Christophe Clavius’ Treatise on the Sphere, published in 1570, had nineteen reprints during the next fifty years. Melanchthon’s textbook, Doctrines of Physics, which was published six years after Copernicus’ book and which attempted to refute Copernicus’ theories, was reprinted nine times before the Revolutions was reprinted a single time (1566); and had a further eight editions later on. Kasper Peucer’s textbook on astronomy, published in 1551, was reprinted six times in the next forty years. The works just mentioned, plus Ptolemy’s Almagest and Peurbach’s Planetary Theory reached altogether about a hundred reprints in Germany till the end of the sixteenth century – the Book of Revolutions, one.
I have two comments on Koestler’s argument. Firstly, Gingerich estimates that both the first and second editions of De revolutionibus numbered between five and six hundred copies, not the thousand that Koestler quotes. Secondly, I think he is to a large extent comparing apples with pears. Most of the books he lists are university textbooks consumed by generations of students. De revolutionibus is a complex original work on mathematical astronomy, which is only really accessible to a relatively small specialist readership, who following Gingerich’s survey appear to have mostly read it.
Koestler’s second justification is much more pertinent and concerns “modern” experts on astronomy. These experts all claimed in writing that De revolutionibus was geometrically simpler that the Almagest. Making this claim based on Copernicus’ arguments in the Commentariolus, where he states that he will need only thirty-four epicycles, whereas Ptolemy needed eighty. The former astronomer Royal, Sir Harold Spencer Jones (1890–1960) makes this claim in Chamber’s Encyclopaedia and Herbert Dingle (1890–1978) repeats it in his Copernicus Memorial Address to the Royal Astronomical Society in 1943. It gets repeated in E. A Burtt’s the Metaphysical Foundations of Modern Physical Science (1924), Herbert Butterfield’s The Origins of Modern Science (1949), H. T. Pledge’s Science since 1500 (1931), and Charles Singer’s A Short History of Science(1941). Koestler argues that these experts had obviously never read De revolutionibus, because in reality, as he points out, if he has counted correctly, here Copernicus uses forty-nine epicycles, whereas Peuerbach in his geocentric Ptolemaic astronomy, a book from which Copernicus learnt his astronomy, only requires forty. Gingerich argued later that Koestler had in fact miscounted and Copernicus only required forty-five epicycles, but Koestler’s point still stands. Copernicus’ geometrical model of the cosmos was not simpler that the prevailing Ptolemaic one.
One cannot, however, claim that Koestler was the first to notice this because a German professor of philosophy, and I can’t for the life of me remember who, had already pointed it out in a paper published in 1948. Despite this, many people still blithely claim that De revolutionibus was mathematically/geometrically simpler than the Ptolemaic system.
Koestler has one more provocation on hand concerning the standard story of Copernicus and his De revolutionibus. There has been much ink spilt over the centuries over the ad lectorum that Andreas Osiander, as editor, added to the front of De revolutionibus informing the reader that he doesn’t have to regard the contents as factually true but merely treat them as a mathematical hypothesis. It is usually argued that this was done without Copernicus’ knowledge. Koestler proposes the theory that Copernicus did know in advance of the addition and produces some not conclusive evidence for it. Ernst Zinner (1886–1970), the legendary German historian of astronomy and Copernicus expert, whom he consulted on the subject agreed with him. I have never read anybody else who discussed Koestler’s theory.
We now leave Copernicus and proceed to the heart of the book, Koestler’s biography of Johannes Kepler, The Watershed. This was the first substantive biography of Kepler in English and in fact there was only one in German, Max Caspar’s Johannes Kepler, which was first translate in English in 1959, the same year as Koestler published The Sleepwalkers.
First off, Koestler’s biography of Kepler is world class. It is not totally free of errors, for example he repeats the myth that Kepler studied theology, he didn’t, he also implies that Mästlin only imparted Copernican cosmology to a selected few, whereas it is now accepted that he held regular lectures for all students on Copernicus. However, both of these myths were widely accepted at the time that Koestler was writing. Also, because Koestler concentrates on the astronomy and cosmology he comes up short of other aspects of Kepler’s activities. For example, his contributions to optics get mentioned positively in passing but are not handled in any depth. Koestler manages to bring all the contradictions and idiosyncrasies in Kepler’s personality and his endeavours to light, whilst treating him sympathetically. Koestler is particularly interested in showing in detail how Kepler asks the wrong questions then sets about in great detail to ‘prove’ the wrong solutions to his questions, whilst inadvertently laying the scientific foundation of modern cosmology and astronomy. This is exactly why Koestler’s book is titled The Sleepwalkers. Astronomers and cosmologist finding the right answers to scientific questions despite themselves rather than in any systematic way, of whom Kepler is the prime example.
Koestler presents a detailed analysis of the Mysterium Cosmographicum and unlike other who try to dismiss its central thesis as some sort of juvenilia. Koestler, correctly, shows that Kepler stayed true to the concepts of the Mysterium Cosmographicum his whole life, even producing a second annotated edition in the 1620s.
It is with his analysis of the Astronomia nova that Koestler shines. He presents his readers with a step-by-step simplified analysis of all of the diversions, meanderings, blind alleys, errors, restarts, and failures that Kepler describes in his “battle with Mars.” Kepler’s presentation in the Astronomia nova is almost unique in the history of science because he describes in detail the tortuous route he took in his endeavours to determine the orbit of Mars from Tycho’s observational data. Normally scientists just present their results and cover up any blind alley that they might have wandered down to obtain those results. Those results being the first two laws of planetary motion. What Koestler fails to notice, perhaps because he largely ignores Kepler’s correspondence with David Fabricius concerning his research, whilst he was writing the Astronomia nova, is that, as James Voelkel has explained in his excellent Johannes Kepler and the New Astronomy,[3] Kepler’s apparent openness over his efforts is actually a carefully constructed narrative designed to convince the reader that he couldn’t possibly have reached any other conclusion.
Enter Galileo! Galileo appears in Koestler’s book only in so far as he interacts with Kepler, or rather doesn’t as Koestler explains, and how he blundered into conflict with the Catholic Church. Koestler introductory paragraph on Galileo is for 1959 a real humdinger. In 1959, Galileo was still the shinning, crystalline “father of modern science,” who could do no wrong. He was the man cruelly and wrongly persecuted by the ignorant forces of bigoted religion. To criticise him in anyway was almost heretical if you’ll excuse the use of the term. Koestler introduces him thus:
The personality of Galileo, as it emerges from works of popular science, has even less relation to historic fact than Canon Koppernigk’s. In his particular case, however, this is not caused by a benevolent indifference towards the individual as distinct from his achievements, but by more partisan motives. In works with a theological bias, he appears as the nigger in the woodpile; in rationalist mythology, as the Maid of Orleans of Science, the St George who slew the dragon of the Inquisition. It is, therefore, hardly surprising that the fame of his outstanding genius rests mostly on discoveries he never made, and on feats he never performed. Contrary to statements in even recent outlines of science, Galileo did not invent the telescope; nor the microscope; nor the thermometer; nor the pendulum clock. He did not discover the law of inertia; nor the parallelogram of forces or motions, nor the sunspots. He made no contribution to theoretical astronomy; he did not throw down weights from the leaning tower of Pisa, and did not prove the truth of the Copernican system. He was not tortured by the Inquisition, did not languish in is dungeons, did not say ‘eppur si muove’; and he was not a martyr of science.
After a brief biographical introduction Koestler sketches the relationship between Kepler and Galileo from, the latter’s reception of the Mysterium Cosmographicum to the fraught reception of his Sidereus Nuncius. Basically, Kepler was very open, friendly, and supportive of Galileo’s work, whilst Galileo ignored Kepler, except when it was advantages for him not to do so. People talk of the Kepler-Galileo correspondence, however, whereas Kepler wrote numerous letters to Galileo, Galileo wrote only two letters to Kepler in thirteen years. Galileo also ignored Kepler’s work in optics and his Astronomia nova. Later he would completely ignore Kepler’s three laws of planetary motion in his Dialogo.
After a brief account of Kepler’s mother’s witchcraft trial, Koestler now analyses Kepler’s magnum opus, his Harmonice Mundi, not as thoroughly as the earlier two works but similarly negatively regarding Kepler’s strange theories of celestial harmony. He does make a rather strange comment, praising Newton for managing to pick out the three laws from amongst the vast quantities of mystical garbage. The comment is strange because he later briefly discusses the Epitome astronomiae copericanae, which has the three laws as the three laws. His account closes with the final years of Kepler’s life.
The Watershed closes with an account of Galileo’s triumphs with his telescope discoveries followed by the arrogant behaviour that led to his downfall, first in 1616 and then the writing of the Dialogo and his final downfall and trial. Koestler gives a very perceptive analysis of the Dialogo acknowledging the high quality of the writing but strongly criticising the poor quality of the science. He goes as far to ask if Galileo ever actually read De revolutionibus because his presentation of it is so warped. His account of the trial is very good and covers all the salient points. For all those who continue to claim that Galileo was imprisoned or even worse thrown into a dungeon, Koestler has a passage that I’m going to quote in future:
Then he had to surrender formally to the Inquisition, but instead of being put into a cell. He was assigned a five-room flat in the Holy Office itself. Overlooking St Peter’s and the Vatican gardens, with his own personal valet and Niccolini’s major domo to look after his food and wine.
Quite correctly, having trashed the Dialogo and Galileo’s contributions to cosmology, Koestler emphasises that Galileo’s fame rests, or should rest, on the Discorsi and not his other writings.
The final section of the book is titled The Newtonian Synthesis. Having reached a high point with his section on Kepler, Koestler seems to run out of steam and this final section is not very good. According to him The Newtonian Synthesis is the combination of Kepler’s celestial physics, the three laws, and Galileo’s terrestrial physics, the law of falls and projectile motion. This picture is far to simplistic, see my Christmas bog post this year. He waffles and does a lot of handwaving about the origins of the inverse square law of gravity. He also repeats the myth that Newton invented calculus to calculate planetary orbits. Newton didn’t even use calculus to write Principia!
The book closes with a thirty-page epilogue, which I’m not even going to attempt to describe or analyse. I have absolutely no memory of how I felt about it fifty years ago. Reading it now, I found it, quite frankly, totally bizarre. Koestler claims that the Greeks were so far by the time of Euclid that a couple of decades were all that was necessary to develop modern mathematics up to and including analytical geometry. He then goes off on a tangent trying to explain why this didn’t happen. My advice just don’t go there.
So, should one read The Sleepwalkers or not? Yes and no. The sections on Copernicus, Kepler, and Galileo yes, the rest definitely not, or at least not without reading a library full of corrective literature. What would I recommend to read instead? Now we have a problem. I’ll try to explain.
Koestler is a brilliant writer. He was a professional journalist, who became a novelist then the author of non-fiction books over a wide range of topics. He is generally regarded as one of the first rank authors of the twentieth century, winner of many awards and considered as Nobel Prize potential. His novel, Darkness at Noon, is regarded as one of the greatest anti-Stalinist works ever written. The Sleepwalkers is a masterpiece of narrative history written by a master storyteller. Even Owen Gingerich, who hated the book for its portrayals of Copernicus and Galileo. and as explained above ridiculed Koestler’s The Book that Nobody Read, admitted that he found the book immensely inspiring when he first read it. A judgement that I totally share. The Sleepwalkers is truly a delight to read; I have never come across another history of astronomy and/or cosmology that come near it as literature. In fact, I don’t know of many books that has the same historical scope, cosmology from Pythagoras to Newton, as high-class literature as The Sleepwalkers. I own four books that cover the same material and more, are on the whole written on a fairly popular level but are written by professional historians, who are all good writers but not literary greats.
The most obvious candidate is John North’s monumental Cosmos: An Illustrated History of Astronomy and Cosmology.[4] An almost nine-hundred-page tome that covers the subject from the bronze age to black holes. It also covers other cultures and regions, such as China, Japan, India, and South America. This is not a literary work like The Sleepwalkers but a direct account of the history. Despite the almost nine-hundred-pages the sheer amount of material covered means that the space attributed to individual topics is comparatively short-compact, dry, and factual. John North is a very good writer but in comparison on a literary level, The Sleepwalkers is Cinemascope & Technicolor, whereas Cosmos is black and white television. Although, the illustrations are excellent.
My second offer is more for those interested in the technical aspects of the topic, Chris Linton’s From Eudoxus to Einstein: A History of Mathematical Astronomy.[5] This is an excellent book but what’s on the cover accurately describes the content. The book largely stays within the limits of mainstream European astronomy–Eudoxus, Ptolemy, Copernicus, Tycho, Kepler, Galileo, Newton, etc–but there is a thirty-three-page interlude in the middle of a five-hundred-page plus total that covers India, Islamic astronomy, and the European Middle Ages. The mathematical aspects, the central theme of Linton’s book are embedded in enough general historical detail to make the book readable and informative.
My third tome covers roughly the same material as Koestler but is literally conceived as a textbook for university students, which however can be read by the interested layman. It is Todd Timberlake and Paul Wallace, Finding Our Place in the Solar System: The Scientific Story of the Copernican Revolution,[6] which I reviewed when it first appeared, so I won’t go into any detail here.
My final volume would probably be categorised as an encyclopaedia and is the Cambridge Illustrated History of Astronomy,[7] which was edited by Michael Hoskin. Once again what is on the cover is inside the book. Large format, not quite four-hundred-pages, it covers the history of European astronomy from prehistoric times down to the early twentieth century, with chapters also on China and Islam. The authors are a veritable who’s who of the history of astronomy, Clive Ruggles, Christopher Cullen, Owen Gingerich, J. A. Bennett, David Dewhirst and of course Hoskin himself, who co-authored five of the twelve sections and was single author of five others. Given the authors and the publisher, this is of course an excellent mainstream introduction to the topic with lots and lots of beautiful illustrations, but it is not a book you are going to curl up in bed with.
This retrospective review of The Sleepwalkers and the search for potential alternatives is already far too long. For better coverage than Koestler of many of the topics in his book I would recommend turning to single topic books but I’m not going to start listing them here, maybe on another occasion. If any readers have any recommendations they are welcome to list and possibly describe them in the comments.
[1] Owen Gingerich, The Book Nobody Read: Chasing the Revolutions of Nicolaus Copernicus, Walker & Company, New York, 2004
[2] Owen Gingerich, An Annotated Census of Copernicus’ De Revolutionibus (Nuremberg, 1543 and Basel, 1566), Brill, Leiden 2002
[3] James R. Voelkel, Johannes Kepler and the New Astronomy, OUP, 1999.
[4] John North, Cosmos: An Illustrated History of Astronomy and Cosmology, Chicago University Press, 2008
[5] C. M. Linton, From Eudoxus to Einstein: A History of Mathematical Astronomy, CUP, 2004
[6] Todd Timberlake and Paul Wallace, Finding Our Place in the Solar System: The Scientific Story of the Copernican Revolution, CUP, 2019
[7] Cambridge Illustrated History of Astronomy, ed. Michael Hoskin, CUP, 1997.