Actually Understanding Physics

[AdamF]

I'm having the same problem with nearly every physics textbook that I've come across (no matter which branch, no matter how advanced or basic), and maybe somebody can help:

After nearly every sentence or claim given in a book, I find myself asking..."How did somebody figure this out? How would somebody think to figure this out? How is this measured? Where's the evidence for the author's claim? Is the claim always true? How do we know? What does this actually mean, and why should I care?"

Nearly every single graduate or undergraduate textbook fails those tests significantly more often than it passes, and it feels like a handful of people in the world actually understand where science has been and where it's going while nearly everybody else just pushes some symbols around pages and becomes reasonably proficient at repeating what they are told.

Is there a recommended place to go for some resources (books, courses, otherwise) where physics subjects are taught through tracing the discovery process with emphasis on phenomenology and legitimately supporting the assertions made in the attempt to explain the subject?

For example: Feynaman's explanation of infinity.

 

[Dale]

Nearly every single graduate or undergraduate textbook fails those tests significantly more often than it passes,

Of course. The questions you are asking are largely historical questions. You should not expect a science textbook to spend much time on the history, that would be a topic for history textbooks.

Did you ever think of purchasing some history textbooks for your history questions?

where physics subjects are taught through tracing the discovery process with emphasis on phenomenology and legitimately supporting the assertions made in the attempt to explain the subject?

IMO this is a very bad way to teach science. My evidence for that is the confusion engendered by teaching the “thought experiments” of Einstein or the false starts in the development of QM.

 

[Master1022]

Not sure for really advanced levels. However, for the beginning of undergrad, it may be worth checking out some of Walter Lewin's lectures from MIT OCW on youtube. He does do demonstrations and examples in the course to solidify the concept and his explanations really helped me to understand the concepts (for Classical Mechanics and Electromagnetism).

Otherwise, if you want to know the exact proofs that were used, you might want to open up the books in which said physicists published their findings.

 

[fresh_42]

I find Wikipedia helpful here.

E.g. to the keyword "elementary charge" I get to
https://en.wikipedia.org/wiki/Elementary_charge#Experimental_measurements_of_the_elementary_charge
and already within this article, I find
https://en.wikipedia.org/wiki/Oil_drop_experiment
or
https://en.wikipedia.org/wiki/Avogadro_constant#Measurement

And from there you can either find new keywords to search for more detailed information, or look up Wikipedia's references and continue there.

 

[AdamF]

I find Wikipedia helpful here.

E.g. to the keyword "elementary charge" I get to
https://en.wikipedia.org/wiki/Elementary_charge#Experimental_measurements_of_the_elementary_charge
and already within this article, I find
https://en.wikipedia.org/wiki/Oil_drop_experiment
or
https://en.wikipedia.org/wiki/Avogadro_constant#Measurement

And from there you can either find new keywords to search for more detailed information, or look up Wikipedia's references and continue there.

Exactly, thank you, that's what I've been using -- the complexity tends to spins a bit out of control when I follow the nodes, so I was wondering if there was some kind of compendium which explained things in this manner and prioritized the story more coherently -- the Russian textbooks seem to be more along this bent, and I've pretty much decided that I'll need to heavily use the original sources to actually understand anything, which isn't very surprising.

 

[fresh_42]

I'm sure there are indeed history books which cover what you are looking for. Something like Physics From Newton To Schrödinger or similar. I have such a book about mathematics, and many mathematical concepts have been discovered in order to solve physical problems. Nowadays the relation is a bit different, and what you were asking for is more of the experimental part. E.g. I found the most interesting part of the black-body-radiation in Planck's vita rather than in the equation or his experiments. What was really interesting was why he looked at it at all.

 

[Chestermiller]

I couldn't think of anything more boring than Physics taught as a history lesson. Students take physics to learn the physical fundamentals and to get practice in applying these fundamentals to solve actual real world problems, in anticipation of being able to do more advanced versions of the same in their professional careers. There is little enough time to achieve these objectives in physics courses, which are typically chocked full of important concepts to learn.

 

[kith]

I have three recommendations:
1) The Feynman Lectures. You cite one of his explanation as an example of what you are looking for, so chances are that you also like his lectures. Have you already tried them? He builds all of physics from scratch and tries to justify everything as much as possible.
2) Malcolm Longair - Theoretical Concepts in Physics. In a couple of case studies, he follows the thought processes of the scientists who discovered important things.
3) Karoly Simonyi - A Cultural History of Physics. A well-written and complete work which has all the details on the history from ancient times to modern particle physics.

I have ordered the texts in decreasing suitability to learn physics from and in increasing demand of previous knowledge. None of them contains exercises, so you run the risk of only learning about physics instead of really understanding physics. You can't achieve the latter without solving problems.

 

[AdamF]

I have three recommendations:
1) The Feynman Lectures. You cite one of his explanation as an example of what you are looking for, so chances are that you also like his lectures. Have you already tried them? He builds all of physics from scratch and tries to justify everything as much as possible.
2) Malcolm Longair - Theoretical Concepts in Physics. In a couple of case studies, he follows the thought processes of the scientists who discovered important things.
3) Karoly Simonyi - A Cultural History of Physics. A well-written and complete work which has all the details on the history from ancient times to modern particle physics.

I have ordered the texts in decreasing suitability to learn physics from and in increasing demand of previous knowledge. None of them contains exercises, so you run the risk of only learning about physics instead of really understanding physics. You can't achieve the latter without solving problems.

Thanks, Kith.

Will check out the Feynman Lectures.

 

[Drakkith]

After nearly every sentence or claim given in a book, I find myself asking..."How did somebody figure this out? How would somebody think to figure this out? How is this measured? Where's the evidence for the author's claim? Is the claim always true? How do we know? What does this actually mean, and why should I care?"

A physics textbook, and also a physics class, are designed to teach you how to do physics, not to teach you the history of physics or to provide evidence for every single assertion and formula in the text. Similarly, most math books do not cover the history of mathematics or tell you why you should care about every single new equation, formula, or whatnot. These textbooks are already hundreds of pages long, I can't imagine having to lug around a textbook two or three times that size just because the authors wanted to include a bit of history.

Also note that you have to have some amount of trust in the text and the class. All of science should be traceable to reputable sources and evidence, of course, but this is absolutely impossible to put down and codify into a textbook or two designed for students. You'd need an entire library to hold all of this information! You have to trust that even if you don't understand the physics you're being taught, it works.

Some of your other questions will be resolved simply by learning more physics.

 

[AdamF]

-- And to the people who don't understand why I'm asking about these things, if you don't think

A physics textbook, and also a physics class, are designed to teach you how to do physics, not to teach you the history of physics or to provide evidence for every single assertion and formula in the text. Similarly, most math books do not cover the history of mathematics or tell you why you should care about every single new equation, formula, or whatnot. These textbooks are already hundreds of pages long, I can't imagine having to lug around a textbook two or three times that size just because the authors wanted to include a bit of history.

Also note that you have to have some amount of trust in the text and the class. All of science should be traceable to reputable sources and evidence, of course, but this is absolutely impossible to put down and codify into a textbook or two designed for students. You'd need an entire library to hold all of this information! You have to trust that even if you don't understand the physics you're being taught, it works.

Some of your other questions will be resolved simply by learning more physics.

Physics is a branch of science, and by extension, learning how to "do" physics entails learning how to apply the scientific method in the form of making observations, constructing hypotheses, testing these hypotheses through experimental design, and so on.

What the "Problem Sets" in 99% of the resources I referenced are doing in this context is briefly touching upon one of the aforementioned steps in the Scientific Method by making every effort to indoctrinate the reader into believing they need to generate a hypothesis in accord with certain constraints while providing no evidence as to why said reader should believe those constraints exists, and then leaving the reader in the dark as to the rest of the process. (Go take a look at any random Problem Set from your typical textbook at any level, graduate, undergraduate, doesn't matter -- and then compare this to how science is actually performed.)

In fact, I was talking with a Post-Doc from CalTech about all of this, and he basically laughed and said the entire academic system is a complete joke in terms of how far removed it is from what actually takes place in the process of making a legitimate discovery and "doing" the subject that it's pretending to teach.

I actually did it your way and I ended up at Princeton in the Graduate School of Mathematics before I decided that it was a total waste of time and walked away from Academia.

So, no.

 

[Dale]

learning how to "do" physics entails learning how to apply the scientific method

I absolutely support this! But note that is substantially removed from your OP. You can learn to “do” science well with very little history. The history is largely a distraction for both goals.

 

[Drakkith]

Physics is a branch of science, and by extension, learning how to "do" physics entails learning how to apply the scientific method in the form of making observations, constructing hypotheses, testing these hypotheses through experimental design, and so on.

Okay. But you failed to mention the actual 99% of what you do in physics. And that's calculate things and solve problems. Everything from forming a hypothesis to building a testable model to designing and setting up test equipment depends on your ability to calculate things and solve problems. That's why all the problem sets are about making calculations and solving problems.

What the "Problem Sets" in 99% of the resources I referenced are doing in this context is briefly touching upon one of the aforementioned steps in the Scientific Method by making every effort to indoctrinate the reader into believing they need to generate a hypothesis in accord with certain constraints while providing no evidence as to why said reader should believe those constraints exists, and then leaving the reader in the dark as to the rest of the process. (Go take a look at any random Problem Set from your typical textbook at any level, graduate, undergraduate, doesn't matter -- and then compare this to how science is actually performed.)

I don't agree with this at all. In fact, I think this is nothing but a bunch of nonsense strung together by someone with a profound misunderstanding of why such problem sets exist. They exist because real world problems require the skills gained when students do this type of work. These problems, in my experience, require that you bring together several different concepts under a certain topic and apply them properly to solve a problem. Which is exactly something you need to know how to do in the real world.

As for constraints, constraints are always taught to students. Things like conservation of energy and momentum, boundary conditions, and other constraints are of fundamental importance to solving these problems, so I have no idea where you're getting the idea that students are being left in the dark.

As for evidence? The evidence is all around you in the functioning of technology, of buildings, cars, and even your own body.

In fact, I was talking with a Post-Doc from CalTech about all of this, and he basically laughed and said the entire academic system is a complete joke in terms of how far removed it is from what actually takes place in the process of making a legitimate discovery and "doing" the subject that it's pretending to teach.

Then the post-doc is a moron who wouldn't be where he was at without the fundamental skills gained from the exact stuff he is criticizing.

I actually did it your way and I ended up at Princeton in the Graduate School of Mathematics before I decided that it was a total waste of time and walked away from Academia.

So, no.

So what? Every year thousands upon thousands of people don't walk away from their graduate programs and learn math and physics just fine. The problem here has very little to do with physics and has much more to do with you. There is no secret to learning physics. It's not some mystical art that only a privileged few with the right talent have any chance of understanding. You learn physics (and all other science) just like you learn any other subject at school. Read the book, do the homework, ask questions, etc.

I'm sorry but I can't take your complaints seriously. They sound like they come from a whiny freshman who doesn't want to make the effort to learn the material and would rather complain about how physics doesn't make sense and that the academic process is all nonsense anyways. None of the issues you bring forward hold up under even cursory scrutiny in my opinion, you just have wildly unrealistic expectations about how physics should be taught and about how easy it should be.

In short; Stop whining. It's your fault you're not learning physics, not the way its taught or the content of the work.

 

[Ray Vickson]

I'm having the same problem with nearly every physics textbook that I've come across (no matter which branch, no matter how advanced or basic), and maybe somebody can help:

After nearly every sentence or claim given in a book, I find myself asking..."How did somebody figure this out? How would somebody think to figure this out? How is this measured? Where's the evidence for the author's claim? Is the claim always true? How do we know? What does this actually mean, and why should I care?"

Nearly every single graduate or undergraduate textbook fails those tests significantly more often than it passes, and it feels like a handful of people in the world actually understand where science has been and where it's going while nearly everybody else just pushes some symbols around pages and becomes reasonably proficient at repeating what they are told.

Is there a recommended place to go for some resources (books, courses, otherwise) where physics subjects are taught through tracing the discovery process with emphasis on phenomenology and legitimately supporting the assertions made in the attempt to explain the subject?

For example: Feynaman's explanation of infinity.

One of my favorite books along these lines is the two-volume set "History of the Theory of Aether and Electricity", by Sir Edmund Whittaker. Volume I deals with the discoveries explained (including false starts, etc.) for "electrical/optical" material, ranging from early works of the Greeks, up through Descartes, Newton, etc (for optics) up through Maxwell's 19 century work and onward to discovery of the electron, and the work of Lorentz. The discourse is not merely verbal; it show developments of the equations and laws, but written in modern notation; it gives references to original publications, so if you can get over to Cambridge, England and read articles that may be in Latin, you can follow the original authors. Volume II deals with the modern era, including the development of special and general relativity, quantum mechanics, etc. Again, the context is historical (showing the numerous successful and sometimes uncesseful attempts, all done in a mathematical fashion using modern notation). Whittaker downplays the role of Einstein in the development of special relativity, and some have been rather critical towards him for that. However, it is definitely useful to see an accurate accounting of the contributions of others, which are often discounted in modern textbook discussions.

If you google "History of the Theory of Aether and Electricity" you will see that the book is not expensive, and I believe you can even download pdf versions for free.

 

[Dale]

including false starts, etc.

This is the key, I think. A good historical treatment must include the false starts but a good pedagogical treatment should avoid them

 

[anorlunda]

Others have already said that comprehensive histories might be voluminous and boring. But here is one history I found to be very entertaining as well as educational.

An Imaginary Tale: The Story of [the square root of minus one]

 

[Dale]

Others have already said that comprehensive histories might be voluminous and boring. But here is one history I found to be very entertaining as well as educational.

An Imaginary Tale: The Story of [the square root of minus one]

I agree that they can be entertaining and educational. But learning the history of physics is not learning physics. It is a different goal.

If the goal is to learn and be able to use the modern theories of physics then obsessing over the history will be a highly inefficient approach, which is the reason that the books take the approach that the OP is complaining about.

 

[fresh_42]

I agree that they can be entertaining and educational. But learning the history of physics is not learning physics. It is a different goal.

Basically yes, and history is more an interesting entertainment. However, it can help to memorize certain facts. As an example, when I had to repeat the Millikan experiment at school, it wasn't actually the charge of electrons which I remember, it was the concept of terminal velocity and viscosity. So in a way this provided more insight than it has been intended to. And similar may happen if one reads about how certain results have been achieved. Another example is Michelson-Morley which directly leads to SR and their names are sufficient here on PF to quote and everybody knows what we are talking about, or Stern-Gerlach. To some extent the historical experiments seems to be part of the story, not only the formulas.

 

[Dale]

Then the post-doc is a moron who wouldn't be where he was at without the fundamental skills gained from the exact stuff he is criticizing.

In defense of the unnamed post-doc the statement by the OP may be a highly distorted version of his actual position. That is partly why we don’t allow private communications as valid references here. Arguments by proxy are pointless

 

[Dale]

To some extent the historical experiments seems to be part of the story, not only the formulas

I agree, but they cannot be the focus as desired by the OP. Also, the experiments failing to confirm the false starts are an important part of the history, but a distraction for learning the modern theory.

 

[anorlunda]

I agree that they can be entertaining and educational. But learning the history of physics is not learning physics. It is a different goal.

True. But reading some of those entertaining history books might satiate the OP's curiosity and show him how daunting it would be to read one background book for every sentence in his physics textbook.

 

[Drakkith]

In defense of the unnamed post-doc the statement by the OP may be a highly distorted version of his actual position. That is partly why we don’t allow private communications as valid references here. Arguments by proxy are pointless

Too true. My apologies to the unnamed post-doc. I was caught up in the moment.

 

[mathwonk]

As an absolutely clueless physics learner, I can identify with the OP on this. To try to address the questions posed by the OP, I would make two suggestions that have helped me;

1) the book "Thinking Physics" by Lewis Carroll Epstein is very down to Earth and user friendly, for understanding the concepts in an intuitive way;

2) In order to grasp where ideas came from, I strongly recommend reading not textbooks but works actually written by the discoverers of those ideas. There are two reasons for this, first those discoverers actually are in possession of the information you want, namely how was this thought of. Second, they are writing at a time when the ideas have not yet taken hold, and are thus motivated to explain, more than writers may do today, just why they advocate these points of view. E.g. I enjoyed reading works by Maxwell, Einstein, Bohr, Born, Pauli, and of course Feynman, possibly the only textbook author I have read who falls into both categories.

Please excuse my naivete', as a physics outsider, but I submit this for what it is worth to the OP, who seems to be asking how an outsider can get closer to becoming an insider.

 

[Dr Transport]

read some scientific biographies...

 

[vanhees71]

I think a lot has been already said, but here are my personal views on the didactical issues discussed.

It's of course a very individual thing, how to best learn about a subject, and physics offers food for a wide variety of different characters of people.

A very rough distinction is that with time physics has specialized into "experimental" and "theoretical" physics. So the first thing for every physics students may be to find out, whether you are more of the "experimental" or the "theoretical" type. Of course, there's no way to do experimental physics without a big portion of theoretical physics and vice versa. That's why the standard curriculum starts with a good mix of both.

It's also a wrong idea to think, one learns physics by just reading books. To really understand physics, both experimental and theoretical, you have to do it. That's why there are both extended lab sessions in experimental physics letting you really do experiments (often even some historically very important once like Milican's oil-drop experiment to determine the elementary charge or the Stern-Gerlach experiment to demonstrate the quantization of magnetic moments etc) and also plenty of problem sets to solve to really do theoretical physics and not just passively reading a book or listen to a lecture.

Finally some thoughts about "history of science/physics". Here I'm pretty undecided. On the one hand, I think it's nonsensical to start a lecture, say about quantum mechanics, with outdated ideas. Unfortunately this has a long and bad tradition. At high school they usually teach students confusing things of "old quantum theory" like wave-particle dualism, photons as particles, Bohr-Sommerfeld orbits in the atom. None of these early ideas survived more than about 20 years (about 1905-1925; I don't count Planck's early work on black-body radiation (1900), because ironically this most conservative of the early quantum physicists had the right intuition about "quanta" to begin with). It's didactically extremely unwise to teach it to beginners in QT, because it leads to more confusion than good. The worst thing is that you have to forget about these wrong pictures, and the pictures are not only wrong in a quantitative but, even more disastrous for the student, on a qualitative level. QT in its modern form is difficult enough. There's no need to confuse students more than necessary with completely outdated ideas that are only historically important to lead to the modern theory.

On the other hand I consider some knowledge about the historical development of physics as an essential part in understanding, where the current understanding reached by about 400 years of research in physics in the modern sense originates from. It can, in contradiction what I just said in the previous paragraph, indeed help the understanding of modern QT to know about the errorneous early attempts to make sense of the observed failure of classical physics like, e.g., the observation that (a) atoms consisting of charged particles really exist (Einstein's work on thermal fluctuations, particularly Brownian Motion (1905) got finally physicists convinced that matter indeed has a atomistic structure, which was clear to chemists somewhat earlier) and that (b) on the other hand matter is nevertheless stable. Classical physics clearly predicts that both is contradictory since accerated charges (as negatively electrons running somehow around a positvely charged atomic nucleus) radiate off electromagnetic waves and thus in a very short time loose their kinetic energy crashing into the nucleus. To the contrary observation tells us that matter is pretty stable (otherwise we'd not exist to begin with) and that any particle or atom of a specific kind are precisely identical to each other. The development of QT is a paradigmatic example for the development of science, and studying its history shows that it indeed was the often paraphrased "scientific revolution" but also a steady progress of experimental and theoretical work towards the so far best theory of physics we have.

So, my advice is to study a large variety of different kinds of physics books first, getting active in learning, i.e., doing a lot of problems in these books. If possible also do experiments yourself (the mix of methods you automatically get when studying physics at a university anyway). Use modern books and don't bother too much about the history of science. If you have gained some understanding of the modern "facts" science has collected and you start to wonder, where all this knowledge comes from, it's a good advice to read some books about the history of science.

A very good book, which is simply great joy to read, is

https://www.amazon.com/dp/1568813295/?tag=pfamazon01-20

About "old quantum theory" a good idea is to read the imho still best biography about Einstein ever written:

https://www.amazon.com/dp/0192806726/?tag=pfamazon01-20