« first day (4604 days earlier)      last day (623 days later) » 
04:00 - 16:0016:00 - 23:00

04:49
@Mr.Feynman I kind of need both. A non-rigorous but quick way to know how to translate, and then later go back to work out the rigorous. An issue with maths presentation is that they go full general and abstract to the point of being impenetrable to students. That might be what maths people need, and we as physicists cannot insist they change, unless, say, we pay them to (See Feynman on this), but for most of humanity, the impetus for the rigour needs to be motivated.
Otherwise, people cannot see why there was a need for the rigour. And maths people also agree: They would recommend that students go get books of (unintuitive) counterexamples, so that the effects of rigour will be best illustrated.
@Mr.Feynman I think it is only a very tiny minority who would insist that innate talent does not matter at all. The very fact that mental disabilities, handicaps, and so forth, provide rather overwhelming evidence. But it is still a fact that intuition can be trained. Nobody said that everybody can be trained to the same levels.
@Slereah Actually, all publishing was just wild west at the time. Einstein himself lived through the transition between editors publishing anything they liked (and rejecting anything they didn't), to the then completely new, budding, peer review process. There is even a crazy story about that, one that only made sense in light of this being the context. One would have to point out that some mistakes are so interesting that they ought to be publishable too.
@Slereah I had that same feeling. Seriously!
@RyderRude On the contrary. Classical-quantum divide is the most in conflict with actual experimental evidence.
@RyderRude I tried to self-study QFT using a dozen textbooks. Got stuck for 5 years++.
@Secret This is only true if the desired measurement commutes with the free Hamiltonian. Position, for example, never commutes with Hamiltonian, and thus will not be repeatable in the way that we always pretend to claim that it would be.
@ACuriousMind It is as if throwing quantum in front of stuff and things get more hairy!
Actually, this is false for both. Beginner level programming is filled with great introductory stuff. But there is a tremendous chasm between Hello World and GUI programming, let alone concurrency and compiler magic.

This is actually the opposite in physics. The standard physics pathway is incredibly well-thought out, because the earlier generations have thought through how to present students with a gradual path towards understanding the deeper stuff. However, the initial fundamental confusions with basic Newtonian mechanics is often something that self-study will never fix. At least not
@RyderRude Why should it have a purpose, again?
@bolbteppa So hoomans are quantum and the galvanometer is classical?
@bolbteppa Manifestly not the case. Big name physicists have been operating using other interpretations just fine for decades.
@bolbteppa If you claim that everything can be described by a Schrödinger's equation, then it is manifestly not classical.
@Slereah Nobody is saying that two photons have to be bound to have, what should be appropriately called rest energy. Because an electron-positron pair does have rest energy, so that after their annihilation the final state should have that too. Note that one does NOT have to define centre of mass for that.
 
1 hour later…
06:29
@naturallyInconsistent oh yes, one does need both at the end of the day. It's useless to be able to understand something rigorously without being able to understand the essence. Examples are an excellent way to understand things
@Mr.Feynman happy dance for convergence
 
1 hour later…
07:38
@naturallyInconsistent hello. What r some example initial fundamental confusions with Newtonian mechanics?
@naturallyInconsistent it need not. But it may have if u buy spirituality stuff. I was just hypothesizing. But yes, the standard answer is that it has no purpose.
@RyderRude Back when Vertiasium was the new kid on the block, he was covering about how the central difficulty of students is not that they are completely blank slates, but that they are coming in with $F=mv$ and we need to make sure they understand that it is wrong
@RyderRude I would even object to characterising it as a hypothesis. Know that patience is running thin with this one
08:07
@naturallyInconsistent yes, i too had this intuition about Aristotle's laws. One other huge problem is that people dont realise that Newton's second law is pretty much just the definition of force
A better way to teach Newton's idea shud b to just say that the universe is deterministic and follows a second order differential equation. This would cut lots of shit
And then the third law and first law can be taught as an observation about the nature of that differential equation
@RyderRude That will just alienate a tonne of students, not least because they would totally not understand what 2nd order DE means.
Saying F=ma causes a shit-ton of confusion imo
@RyderRude This is an incredibly bad description of the situation. N3L is not at all an experimental fact, but rather is an internal self-consistency relation. N1L is a declaration of war.
@RyderRude I am in full agreement, but that is not excuse for what you are prescribing.
@naturallyInconsistent no, third law is not required for consistency. U can come up with second order diff eqns that do not conserve momentum. Third law is an experimental fact
F=ma is pure BS to a new-learner, becuz when it is introduced, neither F nor m have been defined earlier
08:33
Ironically, advanced physics books teach Newton's idea much better than introductory physics books.
@RyderRude Principia introduced it. Blame the people who overzealously simplified things, making everything much worse
U wud often see $m\frac{d^x}{dt^2}=-\frac{dV}{dx}$ referred to as "Newton's second law" in advanced physics books. While this is still somewhat misleading, but it can be fixed very easily
@RyderRude I am saying that it is a consistency thing for physics reasons. I do not at all think that we should be covering Newtonian mechanics using your 2nd order DE viewpoint, so I am not interested in that way of framing things
This approach is much better than saying F=ma
And it is also more fashionable to cast classical mechanics in terms of first order coupled equations.
08:36
@naturallyInconsistent what do u mean by "it is required for consistency for physics reasons?"
@RyderRude This is where I will have to bring in the viewpoint I am advocating for. What we really should be doing, is to define momentum and so forth. And following Hertz, we should eliminate the teaching of forces. The correct way to express N2L is $\frac{\mathrm d\vec p}{\mathrm dt}=\vec F$ and it connects to, say, potential derived forces as $\vec F=-\vec{\nabla\phi}$, so that we can just have $\dot{\vec p}=-\vec{\nabla\phi}$
@naturallyInconsistent the historical way to teach stuff is often very bad. The ideas have been refined a lot since then.
What we then do, is to strip N2L from the scheme, and always refer to the flow of momentum. The gravitational interaction always imparts $mg$ worth of momentum per second, and if a body is to keep stationary, that momentum must leave the body, say by the surface between body and table.
Einstein, Newton and QM founders were very confused back then about the stuff they had just discovered. They had not things organised the way we have today
So the approach of the founders is usually not great
In that way, the one single transfer of momentum from body to table, appears as on the accounting as a debit on the body, and credit on the table. That is the source of N3L
08:47
@naturallyInconsistent yes, this is good pedagogy. But it is not requirwd for consistency
Becuz $\phi$ need not be translationally invariant
This momentum as currency stuff is an experimental fact
@naturallyInconsistent also, we shud note that not all force lawa can be obtained by gradienting a Scalar potential. This is what i meant by "m\frac{d^2x}{dt^2}=-\frac{dV}{dx}$ is still somewhat misleading"
@RyderRude I am not going to dignify your statements as long as you are throwing them out without first learning about them. Quantum theory is basically the exception that proves the rule that the "first elucidation of theories" is a good way to teach and learn physics. And I say this after having looked at all of the big name ones.
@RyderRude No. It is needed for mathematical self-consistency, because a thing can only be called conserved if this one transfer leads to accounting that is equivalent on both sides of the ledger. The most important organising principle of modern physics is the conservation of energy-momentum, and so I am organising the pedagogy to reflect this fact.
@RyderRude I never assumed it to be.
@RyderRude I only illustrated it that way, and did not assume it so.
@naturallyInconsistent but the conservation of anything is not needed for mathematical consistency. U can hav a translationally non invariant universe. It is all consistent
I'm just saying that translational invariance, or conservation, or "momentum as currency" are all experimental facts
@naturallyInconsistent yes, i think u r right about QM. They had the Von Neumann axioms back in 1920s
And Dirac had also unified Schrodinger and Heisenberg mechanics by then
So, for QM, the early stuff is the best
But tbf if we go back a bit earlier, then u wud get QM as a theory of a wave in positon space, rather than a vector in hilbert space. And the former view is actually bad
The former view is still used in teaching introductory QM
So, on second thoughts, I dont completely agree about QM @naturallyInconsistent
The founders are usually very confused because they have just discovered their theory. It hasnt been explored yet
09:04
@RyderRude That is precisely the opposite. If you follow the historical development of QM, you would know that we would be as confused as the pioneers were. The chronological ordering of the discoveries in QM is particularly bad for understanding.
By "early", I meant 1920s QM, when it was actually founded as a new theory. I dont mean Bohr's model and stuff when people were on the process on discovering QM
But Bohr's model shud b taught imo
And other models too
U shud also look up "Newton's bucket argument". Founders usually think of their theory in terms of metaphysics and philosophy too much @naturallyInconsistent
@RyderRude Both momentum and their conservation cannot be intuited from experiment. You can do a billion experiments and not arrive to this way of summarising all these experiments into this particular form.
@RyderRude Again, something quite often disputed within the professors circle
@RyderRude I am very much aware of that. But it does not appear in Principia, IIRC
@naturallyInconsistent practically figuring it out is another discussion. But one thing is for certain : it is not required for mathematical consistency
What I am trying to say, is that we should all attempt to improve upon the pedagogy, but we need to actually have some empirical evidence that whatever we are doing, is actually a measurable improvement, not just claim that it is this or that.
@RyderRude And again, I am saying it is.
@naturallyInconsistent yes, we shud test and improvise i guess. U too r a professor, right?
U seem to have given lots of thought to pedagogy
I think u said that u r a professor. I may b wrong
09:14
@RyderRude No, I am not. I just happen to have heard of all the noisy arguments that professors have over the teaching details. And I do teach.
@naturallyInconsistent but u have not given me the mathematical inconsistency with it
@naturallyInconsistent oh
I dont teach myself
Becuz i am too confused myself :0
@RyderRude As I was saying, the moment we introduce momentum, the whole point to even pay attention to it is because it is strictly conserved. And something is only conserved if it obeys the local, Maxwellian, kind of transfer dynamics. Because of that, it has to appear in accounting ledgers in pairs. So it is a mathematical consistency relation.
As for N1L, that is just to declare that thinkers of mechanics earlier than Galileo are all plain wrong.
N1L is also a definition for inertial frames. I wonder if Newton thought of it that way. I would expect Newtonian was a bit confused about his laws
Or maybe he wasnt confused at all. It's just that those times were before physics theories started geting axiomatized
If Newton had thought in terms of axioms rather than laws, he would have been less confusing
Newton also opposed Leibnitz's 1/2mv2 as a "measure of motion". He thought it conflicted with momentum
It seems like Newton was just extremely confused becuz of his metaphysics thoughts
you are once again trying to understand something created in a completely different intellectual enviroment through an anachronistically modern lens
Emilie Du chatelet brought back 1/2mv2 as an important thing
09:26
the modern "axiomatic" approach to math and physics is far younger than Newton
But that's what i said :P
@RyderRude I think that that is an anachronism.
@RyderRude Axioms are for maths. Physics start from postulates.
@ACuriousMind but one bad thing is that even today, we introduce physics using "laws" rather than "axioms". It makes people think physics is about metaphysics stuff rather than about modeling
@naturallyInconsistent yes, i meant postulates. Sorry :P
If you want to be fair, you also have to note that Newton's calculus notation was so ass backwards that he hindered English maths and physics for a century.
It is also partly why he could not see that energy is a thing that is not opposed to momentum.
People keep claiming that notation does not matter. It fucking does.
Newton thought of his stuff as laws, which is why he was doing so much metaphysics. He thought momentum was a "quantity of motion", which kinetic energy cudnt be according to him
He probably also understood F=ma using metaphysics
09:33
IIRC the principia did not define it as momentum, but as quantity of motion, yes.
@RyderRude N2L is NOT F=ma
According to Newton, F=ma was a law, given his tendency to do metaphysics. He did not think of it as just a useful definition
If he cud think in terms of definition, he wudnt have thought that mv and 1/2mv2 were conflicting
@RyderRude Newton DID NOT WRITE F=ma. How many times do I need to tell you this?
@RyderRude That is just nowhere near the fact of the matter.
I thought u siad it's from Principia
09:37
I said Principia DID THINGS CORRECTLY and DID NOT write F=ma
He wrote $F=\frac{\mathrm dp}{\mathrm dt}$
Oh. U said "Principia introduced it" when i first said F=ma in this discussion. I think it was just a bad choice of words
It was just a confusion
@naturallyInconsistent this does not make things too much better imo
@RyderRude I am going to ask you for a link, a proof of this.
Yes. It is a link to the msg
@RyderRude That is totally false. In Newton's time it was not established that mass is conserved. And Newton would know how to deal with cases where you deliberately added mass to moving stuff. As long as one's viewpoint is with momentum rather than with mass and accelerations, it is much more obvious how to deal with these.
@RyderRude I see. I meant that Principia defined mass and momentum before N2L.
Oh. I told u it was a confusion :)
@naturallyInconsistent but it still just sounds like a definition of force. $m\frac{d^2x}{dt^2}=-\frac{dV}{dx}$ is the right way to go
Or $\frac{dp}{dt}=-\nabla \phi$ is also alright
Like u said
09:48
That is what Hertz was arguing about. Forces are quite a problem even before the quantum revolution rendered it completely unusable.
@naturallyInconsistent yes. People were extremely mean to Leibnitz becuz he introduced alternative stuff to Newton
Anyway, I need to have dinner. bye for now.
He had a nice calculus notation and he also had kinetic energy
Bye
 
3 hours later…
13:02
Time ordering is defined differently for scalar and spinor fields. As long as we deal with them separately it's fine, but what about situations like the Yukawa theory where we have to time order spinor fields and scalar fields $T\{\overline{\Psi}\Psi\phi\}$? It shouldn't be a problem because scalar fields and spinor fields commute, right?
There is this great comedian. Please check out
The experts on the show are not in on it
13:19
Hi all, i'm planning on writing a math research paper regarding the Fukushima nuclear incident. More specifically, I am looking at how long it takes for the nuclides to decay such that it becomes safe for humans to be around.
This is my plan so far. However, one problem with this is that it lacks complexity. Does anyone have any suggestions of higher level math I could use?
Mad
Mad
13:40
Hello friends
Regarding Excitons in Monolayer Wx_2 substances, a paper by 2015 zhang et al. "Experimental Evidence for Dark Excitons in Monolayer WSe2"
it is mentioned that "Below 80k Excitons mostly recombine through fast relaxation channel
"
can someone explain to me the meaning of " fast relaxation channels"?
Information about this subject seem to be not common due to its new nature.
14:01
Modifying quantum postulates is not the way to go. Becuz we shudnt try to apply patch fixes to a theory
So this is y the non linear Schrodinger eqn stuff will never work
Instead, one shud try to derive QM postulates as a limit of a theory with completely different postulates
What do u guys think??
Tensor dentistries transform via the laws of calculus 🤣
@RyderRude If there was a badge titled "Conversation starter" you'd have 1,000,000x
Maybe there is but it doesn't mean what I mean y'know
I cud get a crackpot badge for my chat stuff :P
Nah :) your ideas aren't CP imo
14:08
Unrelated but is there a crackpot badge on PhySE
Wrong $\neq$ CP
@RyderRude doubtful, this is a no no word afaik
CP is almost non PC
Lol
Is spirituality not seen as crackpot
@RyderRude did ST guys maybe succeed in something like that? Maybe ACM knows
@RyderRude in physics.se? No, just off topic lol
No, ST has a hilbert space and unitary evolution and Born rule
It is all usual QM
Assuming u mean string theory
Yes
14:12
Tho i saw some people claiming that spacetime is an emergent phenomenon that is emergent from the worldsheets
That wud make ST a somewhat different framework than QM
Makes sense actually, QM really seems like "rock bottom" because it is the only theory that really has experiment itself embedded in it -- now that should be a criteria for any replacement
Yes. QM is a framework
ST itself is a framework too. But it is not as grand a framework as QM
I think there r levels of frameworks
I mean measurements, if you'd ask newton "hey dude, how does your theory describe measurements?" He'd either say "huh???" Or go into seclusion for 2 months and invent QM lol
e.g. ST and QFT r frameworks within QM framework
But Standard Model is a specific theory within the QFT framework
I mean, some qm precursor, obviously he won't have enough data for the full qm
14:15
@Amit classical mechanjcs is also a whole framework
This is a great way organise physics
Frameworks ---> Theories
Yes but it is incredible that mighty thinkers like Newton never asked the question, what is the limit of measurements. Or if they did, i'll love a reference. It is also a cute thought experiment to ask, how much of QM could be developed via "gedanken" only, previous to planck etc
Is there an official android app for the stack chat?
It's horrible in the browser
Idk. I also use browser
Watch that enter button lol, right?
@Amit this is becuz classical mechanics is a hidden variable theory. So they never needed to ask that
So they never needed to explicitly talk about measurements as a fundamental part of the theory itself
They cud just pretend the hidden variables were the universe's ontology
And measurement devices just reveal the values of those variables
I don't find that a convincing reason. Surely Newton was aware measurement is physical too. He studied light, and you need to use light even in the crudest measurements to see the objects...
14:21
Yes, i think they probably thought about errors in measurements and accuracy and stuff
But QM has made the ontology so much more interesting. There is some sort of duality. One implication of QM cud b a subject-object divide
Maybe he did mention it in some corner of the Principia, that volume be thick
Maybe I'll look at hsm.se if someone asked about it, intriguing
Ideas that could lead to earlier development of qm
Do u guys think De Brogle was justified in inferring the wave nature of electrons from the particle nature of photons
Or is it a case of an accidentally correct prediciton, but with a wrong justification
@Amit there use to be one, but no more
14:27
e.g. Dirac predicted positrons but his justification was a sea
Light does have a wave particle duality, but the wave part of that duality is a classical wave
For electrons, the wave part of the duality is a probability wave
So was De Brogle justified in drawing a parallel?
Oh yeah I did wonder what happened to the good SE app
@ACuriousMind oh 👍🏻
Shud i make a thread about this on PhySE or History of Science?
I want 2 have a thread about "correct predictions done for the wrong reasons"
Is this on-topic for phySE?
@RyderRude in general, I know that physics.SE never rejects historical questions but rather encourages them to be posted on hsm.se. correct me if im wrong anyone
14:34
Ofc the more non historical content it had, the more physics.se suited it is.
How shud i phrase it tho? I want to ask about De Broglie hypothesis but i also want to ask about other correct predictions done for the wrong reasons
I mean, the essence of the question may be non historical, not sure in this case..
Like Dirac's positron prediction
Ask what about them?
Like what r the examples and how the reasons were wrong and y they accidentally turned out to be correct anyway
14:36
If it is just requesting enumeration of such predictions, it is very hsm.se material
If it is asking for the mathematical/empirical details leading to the wrong prediction, it leans more the physics.se way...
I also want to know, is De Brogle hypothesis an example of this
But De Brogle also got the correct equations this way
Maybe you can cross post :) idk
Do u think De Broglie hypothesis is an example of this @ACuriousMind
14:42
@RyderRude If you do not have a specific question about physics you want to ask, the question will likely be closed as too broad
we generally want questions to at least potentially have unique correct answers instead of being "big list" questions
as for deBroglie's relation: Like most of old quantum theory, it's more of a rule of thumb that works in specific situations than a statement of a physical law.
it is significant today more as a stepping stone towards the Schrödinger formulation of QM than anything else
Yes, it works for very special wavefunctions in case of free particles.
But was de broglie justified in inferring that using the wave particle duality of light?
I don't know what that means
De Broglie noticed that light had a particle nature. So he infers that electrons have a wave nature
sure, and then someone did diffraction experiments with electrons and proved him right
Yeah. It was a correct prediction. But is the analogy with light justified theoretically
14:50
I don't know what that means
I mean that light is a classical wave. Electrons r not classical waves. So there shud b no analogy in the wave particle duality
no, I mean, what does "justified theoretically" mean
deBroglie proposed a genuinely new theory when he made that analogy, there is no claim of having derived it from the classical theory
Means "do both light and electrons exhibit a wave particle duality for the same theoretical reason? "
Perhaps De Broglie just got lucky. Idk
as far as I know, deBroglie did not pretend to have any such reason
the starting point is really just "if light can look like a particle, maybe matter can look like a wave?"
@RyderRude Light is not a classical wave.
@RyderRude The wave part of light, that E field, is a probability wave
14:55
@naturallyInconsistent yes, but when we talk about wave nature of light, we mean the classical wave and classical interference
The fact that there is a particle nature of light means that, even at the time, the "classical wave" is a probability wave.
@ACuriousMind then it's justified i guess. De Broglie did not claim that the math was the same
@naturallyInconsistent no, this is a popular mis-interpretation of QFT
The classical interference pattern is not due to individual photons landing probbailistically on the screen @naturallyInconsistent
@RyderRude If you understand how the thing works, you will not claim that this is a misinterpretation.
This is a very tempting mis interpretation of QFT
@RyderRude I mean you can just read his thesis instead of speculating
14:57
@naturallyInconsistent i know how it works. The classical light is obtained from QFT using coherent states. Classical light is not a bunch of photons landing probbailistically
In fact, there is no position basis in QFT
It is a very tempting mis-interpretation. I was also mis-led by it
We ready for a light =/= plane wave discussion? I think I've got a solid argument.
@RyderRude This very much sounds as if you have misunderstood my (and others') frequent rants about positional wavefunctions
it is true that there is no positional wavefunction for photons
@RyderRude I am literally teaching every step of the way from initially having just E and B fields, to single slit interference, to detecting photons. You are just wrong.
@RyderRude This has nothing to do with the correctness of that statement
this does not mean the interference pattern appearing on a detector isn't due to individual photons, because you can determine "surface wavefunctions", i.e. the probability of a photons to cross a particular surface
@naturallyInconsistent this is also wrong. Becuz position space probability amplitude of photons is not the electric field
The electric field is a wavefunctional or an operator in QFT
15:00
@ACuriousMind I am quite confused on this point. Is there a way to get a probability amplitude for photons? I tried a lot of shit and only got a proportionality
@RyderRude Point to me where I made that assumption?
@ACuriousMind but classical light is not an eigenvector of quantum energy
So it cannot be thought of using photons
@RyderRude I don't know what that has to do with anything
the overall state of the light is a coherent state of photons
you said yourself it's a coherent state!
what do you think a coherent state is?
Yes. U do not need eigenstates
Becuz a photon can be a superposition too
So the classical interference is a bunch of photons probabilistically arriving at the screen?
15:05
that's an annoying interpretational issue
because you can model simple detectors by a bunch of quantum electrons inside a classical background field of infalling light, and get the correct detector response
likewise, you can model this as semi-classical electrons absorbing quantum photons arriving probabilistically
And the latter picture is more fundamental maybe
there's a bunch of stuff by Lamb, who is famously "anti-photon", coming up with such "reversed" explanations of quantum interactions between light and matter, where you end up finding out you don't really need to model the light as being composed of photons, maybe
I remember reading this photon interpretation in a book but i thought it was just "lies to kids"
Becuz i cudnt make sense of it in QFT
I'd say it's a bit more than that. There's at least a bit of experimental evidence:
https://www.science.org/doi/10.1126/science.1202218
I more agree with lamb here. The photon picture of classical light seems very handvawy
15:11
This feels like as good a segue as any. I think single slit diffraction is incompatible with the classical view of light as a plane wave. It leads pretty directly to what we're currently discussing.
The issue with plane waves is the Huygens-Fresnel principle.
https://en.wikipedia.org/wiki/Huygens%E2%80%93Fresnel_principle
@WaveInPlace I was just discussing how I am teaching the whole pathway from classical light to photons. It is a standard textbook thing to derive single slit interference from classical plane waves
@WaveInPlace One does not have to work with it
@naturallyInconsistent, are you talking about using the Huygens-Fresnel principle to calculate regions of constructive/destructive interference?
@WaveInPlace As I was saying it is not necessary to start from Huygens-Fresnel principle, and we even know of theoretical issues with it. The wiki article lists many
Of course, it is very convenient to use it, but its status is of a mathematical trickery, not of fundamental physical importance.
Perhaps I'm retreading some well worn ground then. My issue is with the radial waves it invokes at the slit.
Why is that a problem?
15:21
Because they only seem to appear when the slit is close to the wavelength of light in width. If a coherent beam goes through a wide slit it doesn't diffract much at all.
Also, because the observation screen will be taken to be really far away, the radial decay is essentially unobservable. Easily proved with a direct computation, versus the simpler approximation, and you can plot them out and see that it will basically never matter.
@WaveInPlace Known physics. Can be simply proved.
Is the arrow of internal fermion lines going in the direction of particle-number flow just a convention, right? Weinberg says so but P&S say:
Perhaps I need to do some direct calculations. I was presuming that with a wide slit would spread light out in all directions. Instead, coherence is more or less maintained.
I don't understand if they are stating it as a convention. I mean, if I reversed the internal line momentum (I circled it in black, damn it looks like a loop :P) I would get something consistent too because it would reverse all the $q$ signs in the integral
@Mr.Feynman P&S is correct. But it also is equal to Weinberg—since it will be fixed by the Dirac delta distributions, Weinberg can just do anything
15:27
@naturallyInconsistent So is a reversed arrow for $q$ wrong?
@WaveInPlace If a student is bad at maths but good at programming, I'd jkust suggest they do the dayum simulation, insert only the spherical wave propagation, and show that the simulation changes a plane wave passing through a slit to the half-plane-wave-half-spherical-wave thingy that emerges.
I don't understand it that "must be assigned" is stating a convention or the only factually correct way to go
@Mr.Feynman It would just be done as -q in Weinberg, so, yeah, reversed arrow will jsut be bad.
@Mr.Feynman Careful: P&S are talking about the direction of momentum flow. The arrow on the line is not necessarily the direction of momentum flow - that's the convention part
Also, every junction in QED has an arrow in, an arrow out, and a wave. That immediately fixes all the stuff connected to external legs
15:30
@ACuriousMind This distinction should be relevant only for external legs, shouldn't it? For internal fermion lines we don't distinguish particle and particle
@Mr.Feynman I just said no to this...
@naturallyInconsistent I know you said the opposite orientation is wrong, I'm still working to understand why though
Look, the interaction term is $\bar\psi A\psi$, so it must be one arrow in, one arrow out, and one wave
Two arrows in, is $\psi A \psi$ and two arrows out is $\bar\psi A\bar\psi$, both NOT QED
Oh damn that's true
I have a feeling I don't really understand the problem :P
15:35
@naturallyInconsistent I had only thought about this for fermion external legs and internal bosons
Please also see this answer : physics.stackexchange.com/a/666536 . The photon picture is very controversial. Roger Vadim also says it's wrong
That is one of those things that, when I found out, I immediately had a hatred of the standard texts.
@ACuriousMind As of now I feel I don't understand my problem either
Teach me lies, pedagogical lies
I think classical waves r the way to go. This way, u dont have to worry about some crazy change of basis from the wavefunctional basis to the position basis. And even if u managed to do that, u still have to worry about the technicalities of position space wavefunctions
So the better approach is to just get a classical maxwell wave using Gaussian wavefunctionals
15:38
@naturallyInconsistent, alright, back to the math mines. Thanks for the feedback.
@RyderRude Again, that photons do not have unique positional wavefunctions doesn't mean we should throw out hte notion altogether
note that the approach Roger is calling wrong is thinking about photons as particles with wavefunctions just like electrons are - this is correct
Roger is calling approach 2 as wrong. Look at the approach 2 mentioned in the question text
yes, the question says "each with a probability wave associated with them."
Approach 2 is about photons probabilistically arriving and forming the classical interference
it is wholly correct that this is wrong
15:40
@RyderRude I am not surprised that P.SE would have accepted wrong answers. I am not claiming that the photon has a wavefunction in the same way and/or same interpretation that the electron has, but that a probability wave for the photon exists, just not interpretable as a pure probability.
As an aside, how is a tightly coherent beam maintained according Huygens-Fresnel principle? My intuition just keeps saying that it'll spread out instantly, as though it were emerging from a narrow slit.
@RyderRude Classical anything is doomed.
@WaveInPlace see physics.stackexchange.com/q/74572/50583; your intuition is wrong.
@WaveInPlace Is the beam with a wide cross-sectional area or is it a very tight beam?
Very tight.
15:43
Then it will spread out. It would not be appearing as a beam at all.
Is there then a limit on how focused/narrow lasers can be?
see also diffraction-limited system, no one is claiming that you can have coherent beams with widths small against the wavelength
Only wide cross-sections may look like it is travelling like a beam looks.
Because thinking about it, as @naturallyInconsistent points out, it is the form of the interaction that fixes the direction of the arrow and that's the only correct thing to do for the arrow direction
@ACuriousMind, thanks. That seems like a pretty resounding "yes".
15:44
Now, coming to the momentum flow, which is a different thing
And in this day and age, we can not just buy cheap and extremely powerful lasers, but also watch youtube videos of them being beamed from the hand to a gigantic dot on a wall far away
P&S say that it must be in the same direction as the particle number flow i.e. the same direction as the arrow. Is this part a convention?
Because if I reversed the momentum (not the arrow) it would work fine
@Mr.Feynman Thank you. It is always annoying when students insist that we have to have the waves do that unidirectionality thingy. We keep telling them that the spheical wave CAN give you the illusion of unidirectionality, but they would not listen. I just recently answered a question in P.SE about this and he did not learn either.
@Mr.Feynman the convention part is that the arrow on the line points in the direction of particle number flow. That the arrow has to be "one in, one out" at every vertex does not fix the arrow to be like that, you could choose the arrows to be reversed, i.e. point against particle flow.
@Mr.Feynman The external leg could also have the momentum flow the opposite direction to the arrow, so I think, yes? I kinda think that this is a detail that will always automatically get fixed correctly if you understand the process, so one just needs practice.
15:50
but essentially you can choose all arrows to be reversed and it's all just convention - as long as your rules for translating the diagram into an integral produce the same integrals as everyone else's, it doesn't matter
don't fall into the trap of believing that a Feynman diagram "shows physics" - it just encodes a certain term in the perturbation series
Ok i think the position space wavefunction interpretation may do too. But this also implies that we r really measuring position, instead of electric field
Or we can also have a wavefunctional interpretation where we r actually measuring electric field at each point
Both pictures r purely quantum. But the advantage of the latter is that the electric field measurement is well defined at a single point
But the position measurement r handwavy. They r well-defined upto Compton wavelength resolution
"you could choose the arrows to be reversed, i.e. point against particle flow"

but in that case I would have to reverse *both* (all arrows as you say below) fermion arrows, not just the internal one. I meant that once I already have an "in" external arrow, then the internal arrow will be "out"
@Mr.Feynman correct
On that part I'm fine I guess. Now I'm saying that I agree with Weinberg that the internal line momentum being in the same direction as the particle number flow is a convention too
04:00 - 16:0016:00 - 23:00

« first day (4604 days earlier)      last day (623 days later) »