Global famines have dropped the population to 1 billion people and global civilization has collapsed. In an effort to save future civilization some time, you want to provide some information to kick start civilization's regrowth.

We assume a knowledge and tech level equivalent to Europe in 1800. Any math, general knowledge or tech that we had in 1800, they will have access to. Further, the future reader is fluent in one of the languages that these books are written in.

You have to choose exactly three books on physics

(And only about physics. Other topics will be covered in other questions.) By virtue of a print-on-demand press and a generous internet connection (and minimal scruples about copyright law), you can get your hands on the text and diagrams of most any book/article in existence.

The best book choices will:

  • Give future generations stronger pointers for where to go looking for further knowledge. For example, Newton's Philosophiæ Naturalis Principia Mathematica codified the laws of motion and kicked off significant research into physics. Knowing where to look helps a lot. (Side note, I've tried to read Principia and it's completely impenetrable.)
  • Save them some the trial and error of fumbling around on their own. It's well known that once something has been shown to be possible, it can be rapidly imitated.
  • Need not be all immediately useful. It's just fine to have one book be useful for 50 years then the second book suddenly becomes far more useful. And so on with the third.

Printing off all the physics articles on Wikipedia or arxiv.org/physics won't satisfy because...reasons. Only actual books will satisfy.

Preserving the books is a solved problem so you don't need to worry about it. You're responsible only for picking the three books. These won't be electronic copies as we can't be assured that someone will have access to electronics.

Note to responders: While it's true that three books is arbitrary, the number was chosen as it forces hard choices about which books are really worthy. There are two extremes at play: the utterly mundane, "give them normal undergraduate textbooks" and "compress an entire field down to three books". The first isn't noteworthy, while the second is impossible. Try to push your selection of books further towards the highly comprehensible master-works of the field.

This question is a part of the Three Books series. It will grow to cover many and diverse topics, thus, the fairly narrow scope.

While we can be reasonably sure that these three books will be found together, we can't be sure that they will be found with any other sets of books.

  • $\begingroup$ @kingledion I guess we can assume Ancient Greek level understanding of geometry. There is still some scholarship around. $\endgroup$
    – Green
    Commented Jan 10, 2018 at 2:15
  • 5
    $\begingroup$ As far as I can see this question is opinion based and I'm not entirely sure it fits within the on-topic criteria either. It looks like this is a question to be posed to academics or librarians rather than anything involved in creating a new world. Can you think of a good justification for why your question does fit within world building? $\endgroup$ Commented Jan 10, 2018 at 11:42
  • 6
    $\begingroup$ IMHO this is not a very good question, or a good fit to the site, but OH BOY THESE ANSWERS ARE AWESOME. $\endgroup$
    – SáT
    Commented Jan 10, 2018 at 13:05
  • 1
    $\begingroup$ @bendl the math one is going to be tricky since there's so much and math touches practically everything. Thank you for reminding me. $\endgroup$
    – Green
    Commented Jan 10, 2018 at 20:23
  • 1
    $\begingroup$ I'm not making this an answer (since I only have one book), but Isaac Azimov's Understanding Physics covers everything from Aristotle to Einstein, including relevant equations and how they were derived. It would make a good foundation for two other books to build upon. $\endgroup$
    – Skyler
    Commented Jan 11, 2018 at 14:26

5 Answers 5


The textbook University Physics by Sears, Zemansky, Young, and Freedman is going to be far superior to Isaac Newton's work. For one thing, it's designed as a textbook for people who are not familiar with the current state of the art knowledge. Newton was writing for other people like him, who already knew up through everything before him. So he really just needed to explain and justify his additions in that context. This textbook is designed to be read by people taking physics for the first time.

It's a two volume book, so I take it that it would count as two of the three books for the question. However, it is incredibly comprehensive, including not just classical mechanics but also thermodynamics, electromagnetism, optics, and modern physics. Modern meaning relativity and quantum mechanics are both introduced.

Another major advantage is that it is kept updated. Newton was simply wrong about some things. Preserving Newton would thus lead to people being partially wrong, which would not necessarily be made clear in the other books. That's one reason to choose general relativity over quantum mechanics.

Obviously the ideal third book would be a general statement of what else was known about advanced physics so that people could rebuild quicker. So I might pick a layperson's book rather than a physicist's book. The problem is that physicists would tend to concentrate on one thing where you really want a more general survey.

If you don't need a book per se, consider something like Pasco's Complete Physics Experiments. Because understanding the scientific method is at least as important as understanding physics. Most of what we know about physics was discovered in a few centuries. But it required the scientific method to start that.

Another resource is practical physics.

Or if you must have a book, consider How Experiments End by Peter Galison. That's a science history book that details three experiments. I haven't read it, but it looks like the kind of thing that you'd want.

  • $\begingroup$ I appreciate your perspective on Principia. I found it unreadable, as a non-physicist. $\endgroup$
    – Green
    Commented Jan 10, 2018 at 11:27
  • 21
    $\begingroup$ @Green I found Principia unreadable as a physicist. $\endgroup$
    – DanielSank
    Commented Jan 10, 2018 at 17:30
  • 5
    $\begingroup$ I haven't read this textbook, but +several for the point that modern textbooks will be far more useful than historical sources. When Newton wrote his Principia, he was one of perhaps two or three people in the world who could have come up with those ideas; today, it's expected that the average teenager should be able to understand his work. That's not because we're smarter; it's because we've since found radically better ways to explain calculus and Newtonian mechanics. Arguably, the generations of teachers and writers who did that work have contributed far more than Newton himself. $\endgroup$
    – bobtato
    Commented Jan 11, 2018 at 0:53
  • 1
    $\begingroup$ @DanielSank If you think that there is an extant book written by a physicist that fits that definition, you should post it in an answer. My point is that most books on advanced subjects focus because a general survey is not what is needed in our world. You seem to be accusing me of claiming that a physicist would want to focus on one thing for this project. But I never said that. We're using existing books. So you need to find an existing book that is optimized around being one of three rather than one of thousands. $\endgroup$
    – Brythan
    Commented Jan 11, 2018 at 5:37
  • 3
    $\begingroup$ +1 for suggesting textbooks. This is exactly what they are for. The question isn't about what you would put in a museum (Principia) but what you would use to teach humans about physics (textbooks). $\endgroup$
    – Caleb707
    Commented Jan 11, 2018 at 14:08

I don't get how no one thought of Feynman's lectures. There is 0 chance that someone taking Jackson's book for the first time will ever piece together the Electrodynamics again, not knowing other physics than the one contained in other two seminal books. Feynman's lectures are easier to follow and they have been the inspiration for many fledgling physicists.

I would probably vote for writing 3 such books, just as the Bible was written to help preserve the religion over many generations.

The books have to contain a thorough discussion of the scientific method, the basics of mechanics, astronomy, relativity, thermodynamics, quantum mechanics, and general relativity, etc. They should also discuss incomplete or not sufficiently verified theories and point to the edge of the current human knowledge.

  • 10
    $\begingroup$ Yeah. My immediate reaction was "what else would I add after Feynman?" $\endgroup$ Commented Jan 10, 2018 at 8:37
  • $\begingroup$ Why is this not the accepted answer?̣… $\endgroup$ Commented Jan 10, 2018 at 14:24
  • $\begingroup$ It is funny how one of the first discussions in the books is about what single statement of modern physical knowledge one should preserve for this exact situation OP describes. Spoiler: Feyman says it is the atom hypthesis. $\endgroup$
    – M. Winter
    Commented Jan 10, 2018 at 15:55
  • 3
    $\begingroup$ I dunno. I'm a PhD physicist and I generally find Feynamn's lectures kind of unhelpful. $\endgroup$
    – DanielSank
    Commented Jan 10, 2018 at 17:30
  • $\begingroup$ @DanielSank It might be that they don't fit your learning style. I didn't find Jackson's book helpful though I soldiered through it two semesters. $\endgroup$
    – user9981
    Commented Jan 11, 2018 at 7:40

I like HDE's answer, but for the sake of variety here's some different texts.

Disclaimer: a lot of these are undergraduate level, since that's what I'm used to dealing with. There are likely more thorough graduate level texts that offer a deeper and broader understanding, but I think these give a good idea of some of the topics that I might like to include. Also, I think when restarting physics, it might be more desirable to have a book that piques interest, rather than one focused on rigor and deep understanding. Without further ado, here's my list:

Classical Mechanics by John R. Taylor

This is the main choice on which I disagree with HDE. Principia definitely was a tremendous milestone in history, but I think that it leaves out lots of useful classical mechanics simply because there was only so much that Newton had time to discover. Most notably, it leaves out the Hamiltonian and Lagrangian formulations of classical mechanics, which are extremely powerful for solving problems, and even provide an integral part of the foundation for quantum mechanics and QFT.

I think Taylor's book does a good job of covering these topics while still maintaining motivation and not spending too much time focusing on rigor. After all, if the apocalypse survivors don't have access to mathematical literature, it could be incredibly difficult (and perhaps even impossible!) for them to decipher sections talking about symplectic manifolds and smooth sections of tangent bundles.

It also includes several sections on Newtonian mechanics, continuum mechanics, and orbital mechanics, so a lot of Principia's content is covered. The sections on computing might not be useful post apocalypse, but its brief intro to chaos theory is still a very important topic of contemporary math that it would be good to at least be exposed to. Finally, it has a chapter on special relativity, which would be a useful reference.

Classical Electrodynamics by John David Jackson

I agree fully with HDE on this one. Electrodynamics is far too important to leave out, and Jackson is a good resource. Its vector calculus section would also be a good resource for survivors.

An Introduction to Thermal Physics by Daniel Schroeder

This last one is tough, because while I know mechanics and electrodynamics are necessary, there are lots of other disciplines that would be great to have. I went for statistical mechanics, because most of what we deal with in everyday life are large systems, so statistical mechanics is incredibly important for describing the world around us. In addition, it helps bridge the gap between chemistry and physics, and gives a solid motivation and understanding for otherwise incredibly abstract concepts like entropy.

As for the book itself, I love the way it addresses material. It's not very large, and doesn't contain the most information, that's for sure. But it is very clear to follow and has lots of exercises that, while hand-wavy, hint at deeper concepts and got me very interested to learn more, which I think is the most important part of science. You can have the most thorough book in the world, but if it doesn't inspire anyone to learn, there's not much point.

It also talks a bit about quantum mechanics, and while it doesn't go into any great depth, the applications are much more concrete than those normally discussed in quantum mechanics classes. Incidentally, this approach matches the historical discovery of quantum mechanics much better; people were trying to figure out the spectrum of a black body long before some weirdo decided to launch electrons at a double slit.

And that about does it for my recommendations.

It does feel wrong to leave out general relativity, but special relativity is covered and if you remember enough to scribble down that general relativity says mass warps space time, I'm sure someone could eventually figure it out. Also, I think general relativity, although beautiful, is less relevant to society than quantum mechanics. Schroeder contains a few basic problems on semiconductor physics, which is probably one of the most important scientific topics for modern society.

Note: I'm assuming knowledge of math up to at least basic calculus is preserved, otherwise I might replace the third text with a math one. Principia might be good for this purpose, but its lack of more modern notation would make it hard to use as a reference to the other texts.

  • 1
    $\begingroup$ Good first answer, and welcome to WB.SE! $\endgroup$
    – Shalvenay
    Commented Jan 10, 2018 at 4:13
  • $\begingroup$ Welcome to WorldBuilding! If you have a moment please take the tour and visit the help center to learn more about the site. Looking forward to your contributions. Have fun! $\endgroup$
    – Secespitus
    Commented Jan 10, 2018 at 9:11
  • $\begingroup$ The question has been clarified with additional details. I'm not sure if the changes will invalidate your question, but wanted to give you a heads up in case they did. $\endgroup$
    – Green
    Commented Jan 12, 2018 at 22:54

I've chosen three books which, I see, are somewhat similar to an undergraduate or graduate physics curriculum - the last is a stretch, but that's fine. There is an intentional progression from one to the next. The last two are truly textbooks; the first is not, and so is perhaps not as accessible to a student, but I've chosen it for its breadth and its style (which is, yes, a bit dense). They should give this civilization the tools and inspiration to fill in many of the gaps.

1. Newton's Principia

Actually, the full name of the book is Philosophiae Naturalis Principia Mathematica, as you know, but most people like to be concise. The Principia, in a sense, kick-started a revolution in classical mechanics. It is a treatise on a variety of topics in mechanics, including

  • Newton's laws of motion
  • Newton's law of universal gravitation, which works in most cases this civilization would need it (at first)
  • Some notes on astronomy and planetary motion, including what we now call Kepler's second law and support for a heliocentric model of the Solar System
  • Astronomical data on a variety of bodies - not as good as the ephemerides we have no, but pretty good, for the 17th century
  • Basic hydrodynamics, including viscosity and motion in fluids

The material alone would be a treasure trove in itself, but another part that stands out is Newton's style of writing. I haven't read a whole lot of the Principia, but I know that the way Newton presents concepts is fantastic. He emphasizes how he applies the scientific method - which is necessary if we need to train new generations of scientists - and starts from basic principles.

This book doesn't just teach classical mechanics; it teaches you how to do science. That is priceless.

2. Classical Electrodynamics, by John David Jackson

This one has been used for over 50 years as a graduate text on electrodynamics. It's a bit tricky to get through, yes, but it works. Jackson expects some familiarity with the concepts already, but that's fine. Physics is full of conclusions that are not intuitive or easy to see. These scientists, through these books, have lights to guide them along the way. Things will be much simpler for them than they were for, say, Gauss or Faraday or Maxwell.

The exercises from the book that I've seen are good - tough, but good. As with the rest of the text, the author expects you to work. It's not meant to be easy. Having exercises - which are, of course, absent from the Principia - will train the readers to get ready for the third book.

Introduction to Electrodynamics by D. W. Griffiths comes in a close second here. For undergraduates, it's top by a mile. It's clear and concise. But I do want the book to be a bit challenging, and I do want it to go into detail. So I'll choose Jackson over Griffiths because of how detailed it is.

3. Gravitation, by Misner, Thorne and Wheeler

This book is, I belief, considered one of the best tracts on gneeral relativity. It may be intimidating, but if you're looking for a sort of bible on modern physics, this might be one your best bets.

I've picked this book because, like the Principia, it has value far beyond the mere information it contains. Gravitation is an inspiration, a reason for scientists to explore beyond the realms of immediate practicality and applicability. They need something to look to that will tell them to go for knowledge for knowledge's sake; it well nourish their curiosity. Yes, it will take them a long time to understand it, but Gravitation isn't too inaccessible, once you get past its length.

The book is old - going on four decades now - but a newer edition could be more helpful. Recent experimental developments in general relativity - for instance, gravitational waves - would not be included, obviously, but they might be unnecessary. The essence here is that one of the two pillars of modern physics - general relativity - is encapulsated quite well in a single tome.

But what about . . .

  • . . . special relativity? Well, some of that may be evident from Gravitation. Keep in mind, though, that special relativity has its roots in electromagnetism. Einstein's seminal paper is, after all, On the Electrodynamics of Moving Bodies. We struggled through that in my special relativity course last fall; it's not a huge leap. And some of the basic principles, such as the speed of light, can be derived immediately from Maxwell's equations . . . found in Jackson's book.
  • . . . thermodynamics and statistical mechanics? Leaving these out is hard to defend, but I will say that they may be covered in some of the books chosen for chemistry. I was taught thermodynamics in something of a chemistry-based perspective (and this as a course for physics majors!), so my hope is that the scientists will learn the principles of thermodynamics from those resources. Enough to get them started.
  • . . . optics? I feel like this can be determined empirically - at least, the basic principles. If Newton can inspire them to build a telescope, maybe they'll do some playing around. Plus, there may be something in the Principia about optics. I should check that.
  • . . . quantum mechanics? Yeah . . . about that. I was wondering if you'd ask. Well, listen, I was determined to keep a work of classical mechanics and a book on electricity and magnetism, so that left only one option for modern physics. And I chose general relativity. Maybe you'll laugh at me for saying this, but I'm a little worried that unless the scientists discover it for themselves, they'll reject a work on quantum mechanics as a joke. If you've taken a quantum mechanics course, ask yourself how much you'd believe if you didn't have a trusted person teaching you it.
  • 1
    $\begingroup$ This is an impossible question to answer. What about astronomy? No Kepler, so can you even understand Newton's arguments about planetary motion and gravity? There is a big gap in fluid mechanics (no Bernoulli or Lagrange), no atomic theory is presented, and then as you mentioned no optics, thermodynamics, relativity, and quantum mechanics. A question that leaves answers most unsatisfying. $\endgroup$
    – kingledion
    Commented Jan 10, 2018 at 4:48
  • $\begingroup$ With those books, there's the theory. But it will take a long, long time before those can be turned into something practical. In the meantime, you only got 3 books. Shouldn't at least one of the books contain something practical about reproducing prints so at least an entire nation can study and collaborate in this time? $\endgroup$
    – Mast
    Commented Jan 10, 2018 at 8:44
  • 4
    $\begingroup$ Newton's Principia Mathematica... Hmm. It's written in Latin, so you must obviously include a Latin dictionary and grammar book. It's formulated in a very strange way -- specifically, Newton gives Euclidean geometry demonstrations for formulae which could be immediately derived using calculus, because at the time of publication Euclidean geometry was established and calculus was known only by Newton. It covers only a small part of mechanics. Almost any competent book on mathematical mechanics would be better. $\endgroup$
    – AlexP
    Commented Jan 10, 2018 at 13:00
  • 4
    $\begingroup$ @AlexP Translation is handwaved by the OP. Also, you can't restart physics without also restarting calculus; hence Principia. $\endgroup$
    – kingledion
    Commented Jan 10, 2018 at 13:14
  • 1
    $\begingroup$ @kingledion The way I read the question, math is assumed to be contained in another set of books. $\endgroup$
    – DanielSank
    Commented Jan 11, 2018 at 5:32

Most people seem to be instead answering the question "what books would you use to teach yourself?", or "what books would you use for college physics?". But if we're talking about all of humanity learning from these books, we might as well bust out the most enormous, powerful tomes there are.

1. Modern Classical Physics, Thorne and Blandford

Assuming civilization preserved three other books for mathematics, there's no reason to go all the way back to Newton. That's probably the most inefficient way of doing it, because information is always synthesized and compressed over time.

Kip Thorne is the T in Gravitation by MTW, the 'bible of general relativity' that weighs in at over 1000 pages. Thorne really loves huge books, and his recently released Modern Classical Physics clocks in at over 1500 pages. Reviews on Amazon warn that the book is heavy enough to generate gravitational waves; it easily weighs more than a newborn child.

It disposes of Newtonian mechanics in a few pages and proceeds to cover every field of classical physics, including special and general relativity, statistical mechanics, optics, elasticity, fluid dynamics, and plasma physics.

To put that in perspective, some physicists thought that the field was essentially finished by 1900. Modern Classical Physics covers everything known by 1900, including enough applications to jumpstart civil and mechanical engineering, in just half of its pages. The second half is devoted to some of the most important new fields of physics after 1900, namely relativity and plasma physics. Of course, the other important fields all involve quantum mechanics.

2. Quantum Mechanics, Cohen-Tannoudji

This book also spans over 1500 pages, and it's rarely used for courses because its length is unmanageable, mostly from the piles of special topics. To give a sense of it, the third chapter (of 14) alone has 15 appendices, the first of which is a "reader's guide" to tackling the rest of them.

Cohen-Tannoudji works in the field of atomic, molecular, and optical physics, essentially 'applied quantum mechanics'. His enormous book will let civilization retrace the historical development of quantum mechanics, then get a good part of the way to reconstructing lasers, spectroscopy, and the periodic table. The only thing that's missing is the application of quantum mechanics to fundamental physics.

3. Quantum Field Theory in a Nutshell, Zee

Quantum field theory is the quantum mechanical framework used to describe fundamental physics; it's been called 'the language Nature speaks'. Zee's book gives a whirlwind tour, then launches into a remarkable series of applications. He covers the fundamental theory and then the Standard Model of particle physics rather quickly, then links it to condensed matter physics, explaining phenomena like superconductivity and superfluidity. Combined with the practical knowledge from the first two books, that gets us most of the way to understanding how the entire LHC works. Combined with three books on engineering, we might be basically ready to build it!

Zee concludes with speculative topics like supersymmetry, grand unification, and string theory, and I think that's the best part. With these books, we've not only gone to the edge of current knowledge; we've peeked over it into the unknown.

  • 4
    $\begingroup$ The real question here is why the hell a 200 page introductory economics textbook I was forced to buy for a core class costs more than twice as much as a 1500 page tome on classical mechanics. $\endgroup$ Commented Jan 10, 2018 at 13:41
  • 4
    $\begingroup$ @elduderino Economics is Sith science. It costs more to learn and corrupts your soul. $\endgroup$
    – user9981
    Commented Jan 10, 2018 at 13:46
  • 2
    $\begingroup$ @elduderino Kip Thorne does it for the love of physics! He'd probably sell it for free if he was allowed to. (On the other hand, who's richer, Kip Thorne or the average econ prof?) $\endgroup$
    – knzhou
    Commented Jan 10, 2018 at 13:53
  • 1
    $\begingroup$ @elduderino, understanding that should be part of the economics course surely? $\endgroup$
    – Separatrix
    Commented Jan 11, 2018 at 9:33
  • 1
    $\begingroup$ @Green On the contrary, I think the new info makes my 3 choices completely ideal! I chose them at a pretty high level implicitly assuming 1700s-1800s level math knowledge. $\endgroup$
    – knzhou
    Commented Jan 12, 2018 at 22:58

Not the answer you're looking for? Browse other questions tagged .