5 ms·

These kinds of comments block me from doing a physics dive; I don’t know where to start/am worried about starting from a bad point. I don’t know whether to foll

These kinds of comments block me from doing a physics dive; I don’t know where to start/am worried about starting from a bad point. I don’t know whether to follow some of the trails the linked author has laid out if I want to learn more, whether you’re nitpicking, or whether you’re advocating for a much better approach that starts with the details you say this obfuscates. When I was younger and had more time I’d be willing to eat the inefficiency, and the bigger reason for not doing a deep dive is the difficulty, my own laziness, and lack of practical need, so take what I’m saying with a grain of salt, but the time needed to parse criticism like this to see whats legitimate and whats not drives me a bit nuts. I’ve done enough self learning to know that there is no perfect starting point/backtracking is inevitable, but something about the modern intellectual landscape feels very noisy/jargony and particularly prone to inefficient learning. When you say “there is translation invariance that using the Noether's theorem implies the conservation of momentum”, I have no idea what that means. It sounds like a description of a turbo encabulator. I’m sure its not, and that your description is much more efficient than spelling it out from the ground level, but it annoys me how difficult it is to distinguish between a legitimate domain expert jargon meant to condense complexity and unnecessarily obfuscatory jargon that works more like a status/club membership symbol. Maybe its always been like this somewhat and that problem is inevitable, idk.

posterboy 5y ago
Physics is most illustrative in experimentation and good enough approximations. None of the above plays a role in undergrad as far as I can tell, unless perhaps if going for ivy league. There is a very good starting point following the historic developments, maybe iron age stuff, a bit of platonic philosophy, including logic, rhethorics and philology on an a-level level. purely theoretical physics is rather math heavy final year, post grad stuff.
- hither_shores 5y ago
  The linked page is crankery, but most of the physics it's vaguely gesturing at is standard undergrad material. QFT is not, but it's not too far out of reach either.
BrandoElFollito 5y ago
I have a PhD in physics and left academia 25 years ago. I would divide the learning of physics in four categories: - the initial discovery part where you read about something for the first time (nutation, baryons, ...). It is best to have a shallow but interesting approach then, with some (or lots of) liberty on the precision. - the "things are related" part which was extraordinary in my case. It is the part (about 2 or 3 years into the physics curriculum) where you discover that some things are closely related and you can reuse bits and pieces of what you knew from the previous part to see the "big picture". That part is the most fruitful because you can have your own internal discussions about physics and they are not difficult to confirm or not. - the "let's dive into the details". That one is tough, very tough. You have a lot of details where you can miss what you are actually learning. There is a lot of math involved, up to voodoo math such as renormalisation. This is for the ones who want to make physics the main topic of their life (= academia). - the "realization" part when you make peace with a lot of things you learned, where you trust math to drive some parts of physics, where you finally understand that there are some things that you will not understand (or that are not understandable with the current knowledge). This is a phase you get to sometimes after not having done physics for some time. I would warm recommend to spend some time on part one with some "general public" books to get a hint of what is awaiting you. Then to go to step two with a introductory/mid-level book for students of physics, in the "introduction to physics" part. And then look further to selected areas if you want. Physics is marvelous.
gus_massa 5y ago
> how difficult it is to distinguish between a legitimate domain expert jargon meant to condense complexity and unnecessarily obfuscatory jargon Yep. It's a problem at all levels. Sometimes it's also difficult to distinguish real science from well written crackotery You can look at the webpages of a few universities. Many of the course have the list of official bibliography visible, and that's a good start. Try to follow the same order of courses, it's impossible to understand quantum mechanics without a good base of classic mechanics. Sometimes the main book for a physic degree is too technical. You can also try to read the Schaum's Outline book for the topic. They have a lot of examples and exercise, but the theoretical part is shorter. I like them as a side book, but if you don't want to become a super expert, they are fine. If you want books without math, that's a problem. Some are good, some are bad, and it's difficult to distinguish. There are a lot of fun topics that you can learn without too much math. For example there are a lot of things you can learn about particle physics imagining that quarks are just small balls, but some technical details are too difficult without math (for example why there a 8 gluons instead of 9). There are some good collections of divulgation of science that don't have too much math and are checked by a good editorial team. > Noether's theorem The main idea is that if you magically teletransport everything in the universe one mile to the right, then nobody will notice the teletransport. It's important that it's true if you choose any other direction (what does "to the right" mean?) or any other distances (like 1 feet, 1 light year, because "1 mile" is no special). Obviously nobody has tried this experiment, but we as far as we know the laws of physics are the same everywhere, for example the mass of one proton is like 2000 bigger than the mass of an electron here and the mass of one proton is like 2000 bigger than the mass of an electron in a lab on the other side of the earth and the mass of one proton is like 2000 bigger than the mass of an electron in Andromeda. So this though experiment is just a good guess (if you ignore the curvature of the universe due to General Relativity). But if this guess is correct, then the Noether's theorem says that momentum is conserved, that is a property that is verified in a lot of experiments. So it initially looks like a abstract magical though experiment, but the consequence is that there is an important number that is a constant in each real experiment. This constant numbers sometimes simplify some calculations a lot. It's similar to the conservation of energy that in some cases is useful to prove that something is impossible or get the final result without looking at the nasty details of the experiment. The magical translation of everything in the universe is a symmetry, and the idea is that you can discover other symmetries of the universe. Ignoring some technical details, then you can use the Noether's theorem to discover new numbers that are constant in all experiments. Sometimes the symmetry is a symmetry of all the universe that is easy to see, sometimes it's a more abstract symmetry, sometimes is just a symmetry of the experiment and you ignore the rest of the universe. There are many applications of the Noether's theorem.
whatshisface 5y ago
>I don’t know where to start/am worried about starting from a bad point. Start from the basics. You will get plenty of ability to analyze real-world problems from learning Newton's laws. Lagrangian mechanics rarely even applies in engineering contexts because it can't do friction, air resistance or other dissipative terms.
- hither_shores 5y ago
  Lagrangian mechanics can handle dissipative systems. In many cases, friction included, it's as simple as adding a dissipative term to the Euler-Lagrange equations. This seriously complicates the process of deriving them, but physicists were always handwaving their way through it anyway. Hamiltonian mechanics faces more serious obstacles, though if you really want to you can engineer a noncanonical symplectic structure to capture the dissipation. Wouldn't recommend it though.
pas 5y ago
Start from conservation laws, get a feel for the typical mechanics problems. Acceleration, forces, work, energy, momentum. (At first don't try to understand WTF all that has to do with symmetries/invariants.) Then a bit of EM. (Circuit analysis, RLC circuits.) Eventually you'll need some undergrad level math. To get a few very concrete (hehe) examples maybe look into a bit of structural dynamics (resonant frequency, tuned mass damper, mass distribution described by a matrix). Then eventually you need to be able to solve simple differential equations and know what Fourier has to do with radios, and how mechanics depends on calculus of variations. The more examples you work through the more you'll know what you want to learn about. My recommendation is don't try to deep dive into one topic, instead start to look for similarities in other topics, basically do a breadth-first search.
analog31 5y ago
One thing about the traditional sequence of physics topics, is that it progresses (more or less) in parallel with the traditional math sequence. So you can get yourself up to speed on both simultaneously. Also, you don't need to get all the way to the whiz-bang stuff, for it to be both enjoyable and useful. Granted, my degree was 30+ years ago, but today I specialize in designing measurement equipment. If all you end up learning is calculus plus how some mundane things work like circuits and optics, it would hardly be a loss. I don't even claim to understand fundamental stuff like field theory. There's also a problem with explaining physics: Most of us who claim to understand it (I have a physics degree) have never attempted to explain it or even think about it without math. There's the commonly shared anecdote about physicists trying to grasp the "non calculus based" college physics course, and finding it too complex and baffling to proceed.
amluto 5y ago
If you want to dive in to physics on your own, a textbook can be a decent starting point. Griffiths is pretty good, although if you’re starting from zero, it’ll be a challenge.
eutropia 5y ago
If you can find a copy of “Understanding Physics” by Isaac Asimov, that is an excellent place to start.
raydiatian 5y ago
Step 1. Download a university physics textbook Step 2. Read said textbook Step 3. Refer to internet to give relief and contrast to your mental model. Step 4. Take notes Step 5. When you finish that textbook, use your knowledge as a foundation to select and consume the next. Repeat from step 1. There’s a self-learning approach absent of any dependence on ycombinator comments. And it’s plenty efficient—there aren’t even exams if you don’t want to take them!
ephimetheus 5y ago
I know that is not really what your comment is about, but symmetries, conserved quantities and Noether’s theorem are some of the neatest bits of fundamental physics. Basically you observe that stuff stays the same if you move around (not true on earth, but like, in space, generally speaking). The physics is invariant under translation. That’s what we call a symmetry. Now Emmy Noether came up with a relationship and formulae that connects each such fundamental symmetry with a conserved quantity. For translation the quantity turns out to be momentum. Conservation of momentum is one of the most fundamental building blocks of classical mechanics. And we can derive that it is in fact conserved from something as simple as a symmetry. And it doesn’t stop there: next in line: invariance under rotation: angular momentum conservation, which is equally important as momentum conservation. If angular momentum wasn’t conserved, our universe would look entirely different. Ages later people figured out that this symmetry business goes way deeper than initially thought. These U(1), SU(2) and SU(3) from the posted link are notation for other, less intuitive, symmetries. Here they don’t apply to space itself anymore but to quantum fields. With that and Noethers theorem (and some admittedly kind of involved maths) you can derive theories describing all electromagnetic and nuclear interactions. So basically, these symmetries are effectively at the heart of the world (and physics) and I think that’s amazing.

posterboy 5y ago

Physics is most illustrative in experimentation and good enough approximations. None of the above plays a role in undergrad as far as I can tell, unless perhaps if going for ivy league. There is a very good starting point following the historic developments, maybe iron age stuff, a bit of platonic philosophy, including logic, rhethorics and philology on an a-level level. purely theoretical physics is rather math heavy final year, post grad stuff.

hither_shores 5y ago
The linked page is crankery, but most of the physics it's vaguely gesturing at is standard undergrad material. QFT is not, but it's not too far out of reach either.

BrandoElFollito 5y ago

I have a PhD in physics and left academia 25 years ago. I would divide the learning of physics in four categories: - the initial discovery part where you read about something for the first time (nutation, baryons, ...). It is best to have a shallow but interesting approach then, with some (or lots of) liberty on the precision. - the "things are related" part which was extraordinary in my case. It is the part (about 2 or 3 years into the physics curriculum) where you discover that some things are closely related and you can reuse bits and pieces of what you knew from the previous part to see the "big picture". That part is the most fruitful because you can have your own internal discussions about physics and they are not difficult to confirm or not. - the "let's dive into the details". That one is tough, very tough. You have a lot of details where you can miss what you are actually learning. There is a lot of math involved, up to voodoo math such as renormalisation. This is for the ones who want to make physics the main topic of their life (= academia). - the "realization" part when you make peace with a lot of things you learned, where you trust math to drive some parts of physics, where you finally understand that there are some things that you will not understand (or that are not understandable with the current knowledge). This is a phase you get to sometimes after not having done physics for some time. I would warm recommend to spend some time on part one with some "general public" books to get a hint of what is awaiting you. Then to go to step two with a introductory/mid-level book for students of physics, in the "introduction to physics" part. And then look further to selected areas if you want. Physics is marvelous.

gus_massa 5y ago

> how difficult it is to distinguish between a legitimate domain expert jargon meant to condense complexity and unnecessarily obfuscatory jargon Yep. It's a problem at all levels. Sometimes it's also difficult to distinguish real science from well written crackotery You can look at the webpages of a few universities. Many of the course have the list of official bibliography visible, and that's a good start. Try to follow the same order of courses, it's impossible to understand quantum mechanics without a good base of classic mechanics. Sometimes the main book for a physic degree is too technical. You can also try to read the Schaum's Outline book for the topic. They have a lot of examples and exercise, but the theoretical part is shorter. I like them as a side book, but if you don't want to become a super expert, they are fine. If you want books without math, that's a problem. Some are good, some are bad, and it's difficult to distinguish. There are a lot of fun topics that you can learn without too much math. For example there are a lot of things you can learn about particle physics imagining that quarks are just small balls, but some technical details are too difficult without math (for example why there a 8 gluons instead of 9). There are some good collections of divulgation of science that don't have too much math and are checked by a good editorial team. > Noether's theorem The main idea is that if you magically teletransport everything in the universe one mile to the right, then nobody will notice the teletransport. It's important that it's true if you choose any other direction (what does "to the right" mean?) or any other distances (like 1 feet, 1 light year, because "1 mile" is no special). Obviously nobody has tried this experiment, but we as far as we know the laws of physics are the same everywhere, for example the mass of one proton is like 2000 bigger than the mass of an electron here and the mass of one proton is like 2000 bigger than the mass of an electron in a lab on the other side of the earth and the mass of one proton is like 2000 bigger than the mass of an electron in Andromeda. So this though experiment is just a good guess (if you ignore the curvature of the universe due to General Relativity). But if this guess is correct, then the Noether's theorem says that momentum is conserved, that is a property that is verified in a lot of experiments. So it initially looks like a abstract magical though experiment, but the consequence is that there is an important number that is a constant in each real experiment. This constant numbers sometimes simplify some calculations a lot. It's similar to the conservation of energy that in some cases is useful to prove that something is impossible or get the final result without looking at the nasty details of the experiment. The magical translation of everything in the universe is a symmetry, and the idea is that you can discover other symmetries of the universe. Ignoring some technical details, then you can use the Noether's theorem to discover new numbers that are constant in all experiments. Sometimes the symmetry is a symmetry of all the universe that is easy to see, sometimes it's a more abstract symmetry, sometimes is just a symmetry of the experiment and you ignore the rest of the universe. There are many applications of the Noether's theorem.

whatshisface 5y ago

>I don’t know where to start/am worried about starting from a bad point. Start from the basics. You will get plenty of ability to analyze real-world problems from learning Newton's laws. Lagrangian mechanics rarely even applies in engineering contexts because it can't do friction, air resistance or other dissipative terms.

hither_shores 5y ago
Lagrangian mechanics can handle dissipative systems. In many cases, friction included, it's as simple as adding a dissipative term to the Euler-Lagrange equations. This seriously complicates the process of deriving them, but physicists were always handwaving their way through it anyway. Hamiltonian mechanics faces more serious obstacles, though if you really want to you can engineer a noncanonical symplectic structure to capture the dissipation. Wouldn't recommend it though.

pas 5y ago

Start from conservation laws, get a feel for the typical mechanics problems. Acceleration, forces, work, energy, momentum. (At first don't try to understand WTF all that has to do with symmetries/invariants.) Then a bit of EM. (Circuit analysis, RLC circuits.) Eventually you'll need some undergrad level math. To get a few very concrete (hehe) examples maybe look into a bit of structural dynamics (resonant frequency, tuned mass damper, mass distribution described by a matrix). Then eventually you need to be able to solve simple differential equations and know what Fourier has to do with radios, and how mechanics depends on calculus of variations. The more examples you work through the more you'll know what you want to learn about. My recommendation is don't try to deep dive into one topic, instead start to look for similarities in other topics, basically do a breadth-first search.

analog31 5y ago

One thing about the traditional sequence of physics topics, is that it progresses (more or less) in parallel with the traditional math sequence. So you can get yourself up to speed on both simultaneously. Also, you don't need to get all the way to the whiz-bang stuff, for it to be both enjoyable and useful. Granted, my degree was 30+ years ago, but today I specialize in designing measurement equipment. If all you end up learning is calculus plus how some mundane things work like circuits and optics, it would hardly be a loss. I don't even claim to understand fundamental stuff like field theory. There's also a problem with explaining physics: Most of us who claim to understand it (I have a physics degree) have never attempted to explain it or even think about it without math. There's the commonly shared anecdote about physicists trying to grasp the "non calculus based" college physics course, and finding it too complex and baffling to proceed.

amluto 5y ago

If you want to dive in to physics on your own, a textbook can be a decent starting point. Griffiths is pretty good, although if you’re starting from zero, it’ll be a challenge.

eutropia 5y ago

If you can find a copy of “Understanding Physics” by Isaac Asimov, that is an excellent place to start.

raydiatian 5y ago

Step 1. Download a university physics textbook Step 2. Read said textbook Step 3. Refer to internet to give relief and contrast to your mental model. Step 4. Take notes Step 5. When you finish that textbook, use your knowledge as a foundation to select and consume the next. Repeat from step 1. There’s a self-learning approach absent of any dependence on ycombinator comments. And it’s plenty efficient—there aren’t even exams if you don’t want to take them!

ephimetheus 5y ago

I know that is not really what your comment is about, but symmetries, conserved quantities and Noether’s theorem are some of the neatest bits of fundamental physics. Basically you observe that stuff stays the same if you move around (not true on earth, but like, in space, generally speaking). The physics is invariant under translation. That’s what we call a symmetry. Now Emmy Noether came up with a relationship and formulae that connects each such fundamental symmetry with a conserved quantity. For translation the quantity turns out to be momentum. Conservation of momentum is one of the most fundamental building blocks of classical mechanics. And we can derive that it is in fact conserved from something as simple as a symmetry. And it doesn’t stop there: next in line: invariance under rotation: angular momentum conservation, which is equally important as momentum conservation. If angular momentum wasn’t conserved, our universe would look entirely different. Ages later people figured out that this symmetry business goes way deeper than initially thought. These U(1), SU(2) and SU(3) from the posted link are notation for other, less intuitive, symmetries. Here they don’t apply to space itself anymore but to quantum fields. With that and Noethers theorem (and some admittedly kind of involved maths) you can derive theories describing all electromagnetic and nuclear interactions. So basically, these symmetries are effectively at the heart of the world (and physics) and I think that’s amazing.