Help me teach myself Physics via the Feynman Lectures
January 17, 2010 8:15 PM   Subscribe

I've decided to nerd out by working through the Feynman Lectures. Help me turn that desire into a workable plan.

The background: For a while now I've been thinking about the idea of grabbing a copy of the Feynman Lectures and working through them; watching excessive amounts of The Big Bang Theory and some recent chance encounters with math (like this question) have me really missing math and rigorous mathematical problem-solving. My first instinct was to just pick up a copy and a Calc book and get cracking; upon googling around a bit, my understanding is that the Lectures, while Really Cool, are best treated not as a core textbook but as a supplement.

Who I am: A few years out of college. I write and debug code and do related tasks all day, all of it very logical but very little of it rigorously mathematical. In my Freshman year of college I had two semesters of Calc, did reasonably well in it, and enjoyed it, but haven't touched it since graduating; I also had a single-semester non-Calc-based Physics course that I liked a lot, but didn't find particularly challenging. Oh, and a semester of somewhat serious Statistics; at least, it was serious enough to be part of my Math co-requisite for my Comp Sci major alongside the Calc.

What I need: Other than any general advice, I'm looking for specific suggestions on Physics texts to grab as a core to work off of alongside the Lectures; I also definitely need resources to relearn my Calc as I go. Free is ideal, but I'm perfectly willing to pick up a used textbook if it's a good one.

Basically, I want to have fun and get back to mathematical thinking and problem-solving, via Physics, which is A) a topic I've always sort of wanted to know more about and B) a lens that, in that past, has made a lot of math much easier to grasp by providing a grounding for what would otherwise be totally abstract problems. Also, frankly, I really want to be able to say "Yes, I'm working through the Feynman Lectures" and mean it. I recognize that I'll definitely hit a wall at some point, but I want to see how far I can go.
posted by Tomorrowful to Science & Nature (22 answers total) 80 users marked this as a favorite
Stewart's Calculus is the best of the books I've encountered as an American undergrad. I don't think it's unrealistic at all to simply work your way through the chapters (roughly in order) doing the practice problems and testing yourself with the evaluations. Once you're firmly grounded in calc, nothing in the Feynman lectures should give you much of a problem.
posted by phrontist at 8:20 PM on January 17, 2010 [1 favorite]

I highly recommend the book 200 puzzling physics problems.
posted by water bear at 8:31 PM on January 17, 2010 [1 favorite]

Surely you're joking, Mr Feynman!

No advice. Just *hearts*
posted by jbenben at 8:56 PM on January 17, 2010

I find Feynman to be kind of unaccessible unless you are already familiar with the material being discussed. Once you know a bit about Gauss' law or Photo-electron effect and stuff like that, it's rather quite opaque.

I enjoyed learning calculus from the ground up using Spivak.

Proof-based. Very math-y.
posted by chicago2penn at 9:15 PM on January 17, 2010 [2 favorites]

I'd pick up a copy of Halliday and Resnick and your old calculus book or any similar replacement. Feynmen assumes you know calculus pretty well, but modern (American) textbooks can't make that assumption, so they generally review any new calculus when they hit it. The calculus book is there for back-up.

Also, just reading it is of little help. To actually learn physics, you have to do lots of problems. The ones in Halliday and Resnick would work, but if you want them tied directly to the Feynmen lectures (which have no problems) there was a book of exercises published by Caltech that was meant to go with them, but it can be hard to find.

Have fun.
posted by overhauser at 9:15 PM on January 17, 2010

and by once, i meant unless... sorry
posted by chicago2penn at 9:16 PM on January 17, 2010

Until I got derailed by other geeky interests, I tried to do the same as you this summer, albeit with a bit more of a background in math than you might have.

Either of the Stewart calc books (CCC or the other one with the cello on the cover) should be easy enough to follow for self-study; unless you took engineering calculus, you probably used one of them in freshman calc. At my undergrad institution there are (were?) specific first-year calc courses for physics and mathematical physics majors and they used one of the Stewart texts as well. If you really want to challenge yourself, get a copy of Spivak — I took math major calc I/II with the Stewart text but working through a bit of the Spivak book was eyeopening.

As for a good calc-based physics intro text, look for a copy of Resnick/Halliday's Fundamentals of Physics. It's another common first-year text, but it's also pretty good for self-study as well.
posted by thisjax at 9:21 PM on January 17, 2010

Response by poster: A clarification: I'm well aware that I won't just be "reading" Feynman; I actually (apparently incorrectly) assumed that the associated exercises would be bundled with the lectures; apparently I'll just work out of the problem sets in whichever 'core' physics book I end up with, since I'm not really finding the exercises supplemental available online for reasonable amounts of money.
posted by Tomorrowful at 9:23 PM on January 17, 2010

Great idea! though I'm not completely sure how Feynman will fit in your plan. I find the lectures more conceptually than mathematically challenging, often presenting unfamiliar (superbly elegant but slightly off-kilter) arguments leading to a familiar result. But a joy to read of course.

Many physics textbooks for college level treat everything and then some, and they don't go very deep into the subject matter. While they may provide a starting point, you will be experiencing a gap between the problems to study in those texts and the mathematical rigor you're looking for. I have Alonso & Finn, which is, I think, an abridged compilation of their series of specialized texts. Not familiar with Resnick & Halliday.
As an alternative, you could start with a text on classical mechanics (e.g. Marion and Thornton, Classical Dynamics of Particles and Systems) and move on to electromagnetics, statistical physics, etc later on? Will have to think about recommendations for those books.
And I like the no-frills approach of Mathematical Methods for Physicists by Arfken.
posted by gijsvs at 4:30 AM on January 18, 2010

Seconding, Thirding and Fourthing the Halliday and Resnick. I did exactly what you did: Picked it up, started with chapter 1 and worked all the way through the book, doing all the problems. (Or maybe just the odds? Don't remember.) My edition had some additional chapters on relativity and some quantum stuff in a separate volume.

The really great and unexpected thing was that it also somehow (re)taught me calculus along the way. In college, I could do calc problems but I didn't really get it. The sample problems in Halliday are so clear that it effectively teaches both subjects at the same time. Now I can actually set up my own calc problems. It also gave me the confidence that I went on to do some other subjects similarly (notably vector calc via Div, Grad, Curl and all that, another very hearty recommedation).

It's also very helpful to have someone you can run to when you get stuck. That happened to me 5 or 6 times throughout the book. Feel free to contact me if you don't have a local person (not that I'm such an expert...).
posted by DU at 8:14 AM on January 18, 2010

Best answer: Like yourself, I am a computer programmer who was bitten by the physics bug after my formal education was over. It happened to me in 2000, when I was 40. And now, ten years later, I am an editor of The Feynman Lectures on Physics, Definitive Edition, and co-author, with Feynman and Leighton, of "Feynman's Tips on Physics, a problem-solving supplement to the Feynman Lectures on Physics." If interested, you can read about how that transformation occurred, here: .

Previous to my studying the Feynman Lectures, I had only a high school physics background. I did have a good background in (advanced) college-level math, but, like you, I hadn't used it in my career as a programmer. So it seems you are pretty much in the same boat I was in ten years ago, in terms of studying physics from The Feynman Lectures.

Over the past decade, I have read many physics books, and also taught physics, privately and at Caltech, to a number of very bright students, and in my personal (possibly biased :-) opinion, there simply is no better book for introductory physics than The Feynman Lectures, IF the student is bright, seriously interested in the subject, willing to put a lot of careful thought into it .... and committed to work hard on solving problems. But there's the rub! The one glaring fault of The Feynman Lectures as a textbook, often pointed out, is its lack of worked out examples and exercises.

In the original course, Feynman only lectured and did not participate in the recitation sections where the students were assigned exercises, given tests, etc. That is why The Feynman Lectures on Physics contains only lectures, and no exercises. However, other people (including Feynman's co-authors Leighton and Sands) did an excellent job leading the recitation sections, developed _a lot_ of excellent exercises (many suggested by Feynman himself), and, as mentioned in comments above, Caltech published exercise books for the Feynman Lectures - one per volume - in the early 1960s. Additionally, in 1969 Robert Leighton and (Caltech prof) Rochus Vogt published an excellent exercise book for Volume I, called "Exercises in Introductory Physics," which is basically an expanded edition of Caltech's Volume I exercises. Some of those exercises (about 80) were recently republished in "Feynman's Tips on Physics." However, all of these exercise books have long been out of print, they are hard to find, and expensive (and, besides that, the Caltech exercise books for Volumes II and III completely lack answers, which makes it impractical to use them for self-study).

Caltech is currently working on a new edition of The Feynman Lectures, which features better typography, cleaner, more accurate figures, and the correction of about 1000 (mostly trivial) errors (which you can find documented at Also planned for inclusion is a new 4th volume that will include all the exercises used to teach Feynman's original course, plus some that were developed later when other people taught the course at Caltech - about 1000 exercises, all together - many published for the first time with answers, plus a sprinkling of worked-out solutions, given as examples. I can't tell you when this new edition is going to be available, because I don't know. We have been working on it now for several years, and are getting close, but there are many unpredictable mitigating factors that could effect its publication date. I would say that, most likely, it will not come out before late 2011... we hope, in time for the 50th anniversary of the first lecture Feynman gave to his undergraduate students, on September 26, 1961.

If you would like to discuss these or related matters you may contact me by email using my contact address given at the Feynman Lectures website, (I will not be following this blog so do not respond to me here if you want me to see your response.)

Michael Gottlieb
Physics Department
California Institute of Technology
posted by codelieb at 9:44 AM on January 18, 2010 [133 favorites]

I'm a physics graduate student. If you've never taken any physics classes, I would *not* start with the Feynman Lectures. They are way way too sophisticated. I talked with a senior physics professor once and he agreed with me. It's written for people who already know physics and want to go back and see the beauty and insight.

Read some other books first. I second the suggestion of Halliday and Resnick, Physics (not the Fundamentals of Physics). If you want to do some really hard mechanics problems, I would suggest Kleppner and Kolenkow, An Introduction to Mechanics. It's the standard text for advanced freshman physics at MIT, Berkeley, other top schools. For electromagnetism, read Purcell, Electricity and Magnetism. It's a beautiful book. Unfortunately, there aren't really any ideal books for statistical mechanics. The best I've found are these notes:
posted by qmechanic at 1:11 PM on January 18, 2010

By the way, you don't need to really know any math to do mechanics problems. Most of the calculus you need is really basic, unless you're trying to calculate moments of inertia.

Also, a friendlier book for mechanics problems is Morin, Introduction to Classical Mechanics. It shows you how to solve problems.
posted by qmechanic at 1:15 PM on January 18, 2010

Best answer: Dear Tomorrowful,

I strongly disagree with what qmechanic writes, that the Feynman Lectures on Physics are "way too sophisticated" for someone who has never taken a physics course. This is a popular misconception about the Feynman Lectures. Here you can read about this misconception and others related to it:

The Feynman Lectures was used very successfully at Caltech, for almost 20 years, as the primary text for the 2-year introductory physics course required of all freshman and sophomore students (not only physics majors). The majority of these students were studying physics for the first time, and were concurrently studying calculus for the first time. Selected chapters from The Feynman Lectures on Physics are still required reading for this course.

Haliday and Resnick is also an excellent textbook, certainly to be recommended.

Not all books appeal to all students. I strongly advise you to read Haliday & Resnick, The Feynman Lectures, and other physics textbooks, so that you can judge for yourself which is most suited to your tastes, interests and abilities.

Best regards,

Michael Gottlieb
Physics Department
California Institute of Technology
posted by codelieb at 7:17 AM on January 19, 2010 [4 favorites]

I'll back up codelieb and mention that we used the Feynman Lectures in my first-year physics undergraduate course at Harvard to cover basic mechanics.

Granted it was an advanced course and in that first year we also used Purcell for Electromagnetism and Taylor and Wheeler's Spacetime Physics to cover special relativity.
By my sophomore year, we were on Goldstein for advanced mechanics as well as doing introductory quantum using Saxon.
By my third year, we were doing Field Theory and particle physics.

A less advanced physics course was using Halliday and Resnick and so I am familiar with the book. All I can say is that the book lacks the beauty of the Feynman lectures. All the information is there but it lacks something that books like Feynman's lectures or Shu's Physical Universe have in abundance.
posted by vacapinta at 4:20 AM on January 20, 2010

I used to work with a guy who took Feynman's course, way back when he was an undergrad at Caltech. He hated and reviled the course. And this is not a dumb guy. (He got into Caltech, after all.) I don't remember the details of his critique, but he found it impenetrable.
posted by Crabby Appleton at 10:43 AM on January 20, 2010

Whatever you do, do not start with QED. Trust me. But definitely get to it, eventually.
posted by Splunge at 3:42 PM on January 20, 2010

I used to work with a guy who took Feynman's course, way back when he was an undergrad at Caltech. He hated and reviled the course. And this is not a dumb guy. (He got into Caltech, after all.) I don't remember the details of his critique, but he found it impenetrable.

I heard that undergrads dropped out of it, but by the end of the course the room was still packed, only with grad students, Feynman's colleagues, and so on. I can imagine a generic undergrad would find it highly alienating.

But I'm really glad Feynman took a year or two away from his research to create a massive and amazing work like that.


It would be cool if Caltech released it into the commons under the terms of the GPL, but I don't think that will happen. A decade from now, all the bibilogeeks will have scanned it, and it will be on the torrent networks regardless of what people want or don't want.

I think Feynman would have liked it that way. I don't mean to insult codelieb by bringing up a contentious subject.


I'd recommend Mary Boas's "Mathematical Methods in the Physical Sciences" over Arfken, which I'd throw at a grad student who could handle it.


Halliday and Resnick is solid. Remember that you learn physics by doing problems, and thinking about derivations. Rote memorization is right out. And remember that there's no such thing as an insoluble physics problem, only one that needs more time, or a break and a revisit some time later.


I think this effort by India is fascinating, and would make it a front page post if I had the time:

This channel provides technical lectures from all seven Indian Institutes of Technology and Indian Institute of Science, Bangalore.

The Video Courses are organised as PLAYLISTS under the following Categories:

1. Core Sciences
2. Civil Engineering
3. Computer Science and Engineering
4. Electrical Engineering
5. Electronics and Communication Engineering
6. Mechanical Engineering

This collaborative work is as amazing, and as thoughtful as Feynman's work. We're lucky there are such people in the world that want to give of themselves to the commons in this way.
posted by sebastienbailard at 6:05 PM on January 20, 2010

The Feynman Lectures was used very successfully at Caltech, for almost 20 years, as the primary text for the 2-year introductory physics course required of all freshman and sophomore students (not only physics majors). The majority of these students were studying physics for the first time, and were concurrently studying calculus for the first time.

On the contrary, Caltech freshmen are required to have taken calculus and physics in high school. I have no reason to believe this is a new development, as the comparable Harvey Mudd College had the same requirements in 1990.
posted by exogenous at 9:48 AM on January 21, 2010

Best answer: You could also just watch them.

(Full disclosure: I made that app.)
posted by endquote at 2:23 PM on January 21, 2010 [5 favorites]

Mod note: few comments removed - silverlight derail needs to go to metatalk or email
posted by jessamyn (staff) at 2:56 PM on January 24, 2010

Mod note: Final update from the OP:
At long last, the new edition of the Lectures mentioned in the Best Answer are now available, and available online for free to boot.
posted by cortex (staff) at 4:52 PM on August 25, 2014

« Older The last few links in the chain.   |   What next, Picasso? Newer »
This thread is closed to new comments.