Is Landau/Lifshitz series suitable for learning?

[bacte2013]

I am interested in learning about the classical mechanics, quantum mechanics, and thermodynamics as my current research in the mathematics and microbiology will involve them. I found Landau/Lifshitz series on Amazon, which seems to cover the main branches of physics.

Unfortunately, I did not take any physics since my AP Physics approximately four years ago; however, I heard that general physics is not really required as introductory books on those physical branches cover the general physics too. My mathematical background is the following: introductory analysis, complex analysis, general topology, linear algebra, calculus, and abstract algebra. I am currently studying real analysis and algebraic topology.

Is l/S series good for a beginner, given that one can allocate a large amount of time for studying? If not, could you recommend me some books?

 

[Mondayman]

The LL series is pretty advanced and not really suitable for a beginner. Halliday-Resnick-Walker Fundamentals of Physics is pretty much the standard introduction. Kleppner-Kolenkow mechanics and Purcell EM are more advanced introductions.

 

[wrobel]

the first volume is not suitable

 

[vanhees71]

I disagree in that statement! All 10 volumes of LL are excellent books. They are amazingly modern although partially about 40 years old. The only volume I find outdated is vol. IV on QED, although also this volume contains some gems not found in other textbook literature but only in original papers.

Of course, LL is not an introductory textbook but an advanced book at the advanced undergraduate and graduate level. It's a tough and a very condensed representation but it's always very much grounded in physics. My favorite volumes are vol. II on ("microscopic") classical electrodynamics, which avoids all the pedagogical balast of Purcell, which in my opinion only hinders the beauty and simplification of the relativistic approach rather than emphasizing it. In LL it's immediately visible from the first page on; and volume X on kinetic theory (transport equations a la Boltz up to non-equilibrium QFT in the Schwinger-Keldysh real-time formalism), also providing the non-nonsense physics approach as all of the 10 volumes.

For the classical-physics part it's only comparable with Sommerfeld, which for me is the best series on theoretical physics ever written. Of course LL is more modern in its approach emphasizing the general symmetry principles much more than Sommerfeld.

On the introductory level, only the Feynman Lectures come to my mind which are comparable in quality to LL!

 

[atyy]

Unfortunately, I did not take any physics since my AP Physics approximately four years ago

If you got an A on AP Physics, you can certainly try LL Volume 1 on mechanics.

 

[QuantumQuest]

To talk in my own experience as a non-physicist with a background in CS, I had a hard time to follow the LL texts once, but they are really worthwhile for every page and bit they contain. I you find it difficult to follow the text, you can try some introductory books mentioned above, but I definitely recommend to study LL at some time if you really want to learn Physics. They are among the best textbooks ever written for Physics.

 

[wrobel]

"Mechanics" by LL contains errors. For example, it is well known that the equations of nonholonomic mechanics do not follow from the variational principle of Hamilton. This fact has serious physical and geometric reasons. For details see for example
[Nonholonomic Mechanics and Control (Interdisciplinary Applied Mathematics), Anthony Bloch, et al]. Nevertheless LL claim that they obtain equations of nonholonomic mechanics from the variational principle.

 

[SuperDaniel]

I am interested in learning about the classical mechanics, quantum mechanics, and thermodynamics as my current research in the mathematics and microbiology will involve them. I found Landau/Lifshitz series on Amazon, which seems to cover the main branches of physics.

Unfortunately, I did not take any physics since my AP Physics approximately four years ago; however, I heard that general physics is not really required as introductory books on those physical branches cover the general physics too. My mathematical background is the following: introductory analysis, complex analysis, general topology, linear algebra, calculus, and abstract algebra. I am currently studying real analysis and algebraic topology.

Is l/S series good for a beginner, given that one can allocate a large amount of time for studying? If not, could you recommend me some books?

Hi Bacte2013,

Let me add that I am in a similar position to yours: I get my electronic engineering degree in 1994, so, I haven't review math since... maybe 1992. Despite of that, I had that project to study physics delayed due to my work. Finally, a couple of years ago I started to study by myself, and, some months later I came across with this forum, and I received a lot of help.

One of the guidelines was to study from L&L which sounded odd at that very beginning. I started to look for the books, and, as one guy told me, I bought the L&L on Amazon and AbeBooks with the following criteria: I always chose the series printed in Britain by Pergamon Press, in all cases, used, and most of the time, expensive. I made that because Pergamon published them at the time Landau was still alive and the translators were reliable, according to the advice I had received. Imagine that I didn't bought them in my maternal language, Castillian (or Spanish as some people call it).

Since then, I confess I did a big effort, specially with Mechanics, the first one: I had to review maths and sit on the desk and study, review, study, review... etc... I did, so, you can do it too!

I do not give advice neither after 25 years of experience in engineering which is my field, and, less in physics, which is not my specialty. The only thing I can tell you is the fact that thank to the suggestions of the guys here, I had the chance to do what I wanted, and from the best sources. Finally, you can find support here: some of the guys are professors, and do not forget that you have to struggle in order to understand.

Good uck with your project.

D.

 

[bacte2013]

I disagree in that statement! All 10 volumes of LL are excellent books. They are amazingly modern although partially about 40 years old. The only volume I find outdated is vol. IV on QED, although also this volume contains some gems not found in other textbook literature but only in original papers.

Of course, LL is not an introductory textbook but an advanced book at the advanced undergraduate and graduate level. It's a tough and a very condensed representation but it's always very much grounded in physics. My favorite volumes are vol. II on ("microscopic") classical electrodynamics, which avoids all the pedagogical balast of Purcell, which in my opinion only hinders the beauty and simplification of the relativistic approach rather than emphasizing it. In LL it's immediately visible from the first page on; and volume X on kinetic theory (transport equations a la Boltz up to non-equilibrium QFT in the Schwinger-Keldysh real-time formalism), also providing the non-nonsense physics approach as all of the 10 volumes.

For the classical-physics part it's only comparable with Sommerfeld, which for me is the best series on theoretical physics ever written. Of course LL is more modern in its approach emphasizing the general symmetry principles much more than Sommerfeld.

On the introductory level, only the Feynman Lectures come to my mind which are comparable in quality to LL!

Thank you for the great advice! Is it okay to read both L/L series and Feynman Lectures together? I do not want to start with general physics as I think books like L/L covers the topics in depth, and I do not have enough time budget to allocate for the general physics. I acquired L/L and Feynman Lectures books from an emeritus professor in physics; he gave them to me before leaving for a long travel. Maybe I should buckle myself and try to read the first volume.

 

[bacte2013]

"Mechanics" by LL contains errors. For example, it is well known that the equations of nonholonomic mechanics do not follow from the variational principle of Hamilton. This fact has serious physical and geometric reasons. For details see for example
[Nonholonomic Mechanics and Control (Interdisciplinary Applied Mathematics), Anthony Bloch, et al]. Nevertheless LL claim that they obtain equations of nonholonomic mechanics from the variational principle.

That sounds very alarming...I am planning to read both L/L and Feynman Lectures, and I hope that Feynman books will help me to clarify some errors in the L/L. As mentioned on my latest post, I basically got L/L series and Feynman Lectures for free, so I am hoping not to purchase extra books.

 

[Amrator]

Hey guys, maybe it's best to create another thread for a discussion like that.

<<Mentor's note: the discussion of nonholonomic mechanics in LL has been moved here.>>

OP, considering your mathematical background, try starting with Taylor's Classical Mechanics.

 

[mpresic]

I do not think the L&L series are suitable to a beginner. Your mathematical background is very well developed. An in-depth study of LL would involve at least an equal amount of effort as that was required to learn the mathematics to this advanced level. I truly think you (anyone) could spend at least a decade or more studying the entire 10 volume series. I think research in microbiology would not require or even desire this extraordinary effort.

Physics Grad students generally use Jackson for EM, Sakurai or Shankar for QM, and Goldstein for CM, and Pathria, Beale for Stat Mech. I think using LL volumes 1 and 2 for CM, EM and Gravitation, Volume 3 for QM, Volume 5 for Statistical Mech, you have a graduate program equal to the one typical of grad students, but the effort required to study LL would be even greater. Vanhees (likes LL) and refers to Sommerfeld (Arnold). I heard Sommerfeld's students felt he was an exceptional lecturer.

I feel Sommerfeld is excellent, more discursive than LL, illuminating, and erudite. I think you may be very happy with Sommerfeld, but you can examine LL to see if you like it better. The entire Sommerfeld series is shorter than LL. It is probably better background for microbiology. I believe the six volumes of Sommerfeld has even been condensed to two volumes, but I do not know how complete they are.

 

[bacte2013]

I do not think the L&L series are suitable to a beginner. Your mathematical background is very well developed. An in-depth study of LL would involve at least an equal amount of effort as that was required to learn the mathematics to this advanced level. I truly think you (anyone) could spend at least a decade or more studying the entire 10 volume series. I think research in microbiology would not require or even desire this extraordinary effort.

Physics Grad students generally use Jackson for EM, Sakurai or Shankar for QM, and Goldstein for CM, and Pathria, Beale for Stat Mech. I think using LL volumes 1 and 2 for CM, EM and Gravitation, Volume 3 for QM, Volume 5 for Statistical Mech, you have a graduate program equal to the one typical of grad students, but the effort required to study LL would be even greater. Vanhees (likes LL) and refers to Sommerfeld (Arnold). I heard Sommerfeld's students felt he was an exceptional lecturer.

I feel Sommerfeld is excellent, more discursive than LL, illuminating, and erudite. I think you may be very happy with Sommerfeld, but you can examine LL to see if you like it better. The entire Sommerfeld series is shorter than LL. It is probably better background for microbiology. I believe the six volumes of Sommerfeld has even been condensed to two volumes, but I do not know how complete they are.

Thank you very much for your insightful advice! My upcoming research built upon my current work in the applications of machine learning and topology to microbes and their evolution, where the next direction is investigating the possible effect of thermodynamic values in the certain main regions of the microbial structures and quantum states of the microbial molecules to microbial evolutionary patterns (I am sorry that I cannot go in detail for now). I think such topic requires statistical mechanics, quantum mechanics, and thermodynamics; however, I thought that they built upon classical mechanics and electromagnetism.

I will check Sommerfeld's books and see if I can understand the exposition, compared to Feynman and L/L. I just started out with L/L and Feynman as I got them free, but I am more than willing to invest money now for other books.

 

[mpresic]

I think electromagnetism will be less helpful than the other areas. The three areas you mention find their support in classical mechanics. I think when it comes to thermodynamics, there is a possibility that the areas of thermodynamics touched on by chemists may be more than the thermodynamics that physicists treat. You will probably discuss your research with chemistry faculty as well as physics, and biology.

I know one poster suggested Landau and Lifshitz got some things wrong. That poster has a strong background in differential equations and mechanics. It is not at all uncommon for experienced researchers to find mistakes, inaccuracies, and unclarity in textbooks from prominent researchers and scientists. I feel even Feynman lectures contains inaccuracies and unclear passages. The mathematics and physical arguments in these textbooks tell to what extent their treatment applies, and it is better to "learn the rules" before "worrying about exceptions", where their textbook treatment is incomplete or inapplicable.

 

[vanhees71]

Here, I strongly disagree. Thermodynamics without quantum statistics is a pretty ununderstandable subject, and you should learn QT, including many-body theory. This is possible but very difficult without a solid understanding of classical electrodynamics.

Also for sure there are typos and maybe even mistakes in any textbook, including LL and Feynman. In LL vol. 1 for sure is a big mistake, concerning unjustified statements about integrability. Contrary to the statement, the integrable systems are the rare ones in classical mechanics. Almost all systems are not integrable.

I only disagree with the statement that nonholonomic constraints are treated erroreneously in LL vol. 1. I think it's a correct derivation (in the usual physicists' handwaving way to handle the mathematics) from the d'Alembert principle in variational form. In this point Goldstein is wrong, promoting "vakonomic dynamics" as the correct equations, although these are known to have serious problems leading to sometimes physically untenable results.

 

[bob012345]

On the introductory level, only the Feynman Lectures come to my mind which are comparable in quality to LL!

I doubt that. Feynman himself alluded to the fact that he was not that interested in teaching undergrads.

 

[guitarphysics]

I completely agree with mpresic. LL is great for someone who wants to become a theoretical physicist. It's great for someone who wants to see how elegant physics can be, and how it can be built. It's also good to see into the mind of a great theoretical physicist. It is not great, however, for someone who wants to learn some physics as support for microbiology research!

It seems like what you'll need most of is quantum mechanics and thermodynamics. Given this, and your mathematical background, I recommend to begin either with Taylor's Classical Mechanics or Goldstein's book for CM. Maybe Goldstein's would be better- I don't like it much but it's because I'm not much of a math guy, but you seem to be. The thing about Goldstein is that it's planned to basically make an easy bridge from CM to QM, so right after that you could study Shankar's PQM, or Griffith's introductory book. I love PQM and don't like Griffith's QM book much, but maybe Griffith's book would be better for you; it gets to the point much faster, and focuses less on the foundational aspects of QM. Can't give advice on thermo, I'm guessing following a standard higher level undergrad course's text would be fine.

If you want to learn this stuff well, it'll take quite a long time, and a lot of effort too. Your math background will ease your way considerably, but I'm guessing it'll take at least a year to get all of this down, and that's if you're dedicating a decent amount of time to this. What I'm trying to say is- you can certainly do this, but be aware of the time and effort you'll have to put into it.

 

[bacte2013]

I completely agree with mpresic. LL is great for someone who wants to become a theoretical physicist. It's great for someone who wants to see how elegant physics can be, and how it can be built. It's also good to see into the mind of a great theoretical physicist. It is not great, however, for someone who wants to learn some physics as support for microbiology research!

It seems like what you'll need most of is quantum mechanics and thermodynamics. Given this, and your mathematical background, I recommend to begin either with Taylor's Classical Mechanics or Goldstein's book for CM. Maybe Goldstein's would be better- I don't like it much but it's because I'm not much of a math guy, but you seem to be. The thing about Goldstein is that it's planned to basically make an easy bridge from CM to QM, so right after that you could study Shankar's PQM, or Griffith's introductory book. I love PQM and don't like Griffith's QM book much, but maybe Griffith's book would be better for you; it gets to the point much faster, and focuses less on the foundational aspects of QM. Can't give advice on thermo, I'm guessing following a standard higher level undergrad course's text would be fine.

If you want to learn this stuff well, it'll take quite a long time, and a lot of effort too. Your math background will ease your way considerably, but I'm guessing it'll take at least a year to get all of this down, and that's if you're dedicating a decent amount of time to this. What I'm trying to say is- you can certainly do this, but be aware of the time and effort you'll have to put into it.

Thank you very much for your advice! I have been reading L/L series but I found that the whole series is very difficult to comprehend...I just checked out Goldstein, and I found the book to be very interesting and readable; someone recommended Marion and Taylor before, but I dislike them.

Is it possible to read Shankar along with Goldstein? Looking from the table of contents, Shankar covers the necessary prerequisites in Chap. 1 and 2, so I thought it might be a good idea to read Shankar and reinforce the prerequisite topics with Goldstein (the mathematical knowledge is already covered). Also it is a good idea to study electromagnetism before going into QM or thermodynamics? If so, could you recommend me a source?

 

[guitarphysics]

No problem :). I think it would be fine to read Shankar along with Goldstein; the first few chapters might even match up at some points (I'm thinking about the Hamiltonian/Lagrangian approach to CM). As for electromagnetism- as someone already pointed out, it seems unlikely that you'll need to know much about EM for what you said, but it's your research, so you obviously know more about it than any of us. If you suspect you'll need to know EM, I recommend going with Griffiths' book (intro to electrodynamics) for that. I doubt you'll need to know electromagnetism before QM or thermodynamics. EM, in my opinion, is the most disjoint physical theory from the rest (of the 4 cornerstone physics subjects), but it still does help to know it when looking at other subjects (e.g. it's a field theory so, for example, LL treat it side-to-side with relativity, or for example in QM when analyzing the hydrogen atom, you use the potential familiar from EM).

 

[smodak]

I am going to suggest a little different set of books than already suggested here and they can be read in parallel. They both are fantastic. I am not sure why are they not used more widely - perhaps the books are a bit newer.

For Classical Mechanics, my favorite book is (Spivak)
https://www.amazon.com/dp/0914098322/?tag=pfamazon01-20
Don't be fooled by the title. It is not written for mathematicians. It has excellent physics insights.

For Quantum Mechanics, my favorite book is (McIntyre)
https://www.amazon.com/dp/0321765796/?tag=pfamazon01-20
I love the spins first approach to quantum mechanics. It is much faster to learn (for me at least) and does not require full classical knowledge and historical introductions.