This is probably the greatest concise explanation of neutrino oscillations I’ve ever heard. I would love to see another video continuing the story onto sterile neutrinos, neutrino-less double beta decay, and leptogenesis!
When a dunce such as myself can understand all of this, you need to know you've done an excellent job with your presentation! Perhaps the new physics will be your discovery! 💝
Honestly I really find this as most underrated Physics Channel on RUclips or anywhere... On thr internet... He posses a quite a grip on language... Explaining everything with Mathematics and Diagram in a way that makes iteasy for anyone to understand and do not get any wrong assumptions regarding the subject/topic which is talked about... All and all Just keep it up ... bro ....👍👍👍
A elegant argument is that, wheter or not the mass eigenstates change, the fact that there is a variation in neutrino flavor immediately implies they experience time, as per relativity, and only massive particles can experience time. I think it's shabby, but it's nice sounding.
I like the fact that you actually went into some detail. Most channels are wouldn't explain anything from this subject and just say, the standard model is not complete because neutrinos have mass. However, I almost didn't anything in this video because of all the math I had never seen before.
The probability is the length squared of that arrow. So the graph at around 10:45 should look different, it should be the square of what's shown. It is a cos function shifted and scalaed to between 0 and 1 given that initial state.
OK - I got lost about 2/3 of the film, but still I enjoyed it very much! Great channel. Regarding neutrino mass it reminds me of another mass that we cannot observe directly - dark matter. Perhaps there is a link there?
@a1nd23 This is a great connection to make! And it has, in fact, been studied. It turns out that the standard model neutrinos (the ones that we *have* observed) are too light to be dark matter. Essentially, in the early universe, they have too much energy to be able to be bound by gravity to form stable galaxies and other large structures. However, it is possible that the sterile neutrinos (the ones we haven't detected) could possibly make up dark matter. In fact, there is a model from Shaposhnikov known as the "neutrino minimal standard model" which extends the known standard model by 3 sterile neutrinos and can explain a whole slew of problems in physics, including dark matter. One big issue with it is that the sterile neutrinos have to be quite massive and have a huge hierarchy (2 neutrinos have to be ~1000 times heavier than the other) in order to actually make the model work out correctly. This is not only aesthetically displeasing, but it also would make them incredibly hard to detect. There also could be some signals from this that we might expect to see, but haven't yet, so that is another issue with it. Either way, it is a possibility, but until anything is detected, we can't know for sure!
Personally I enjoyed the show and subscribed..Based on an incomplete data set of a SINGLE video, the following advice could be unjustified but here it is.. Despite your disclaimer, you are STILL getting overly involved in too much detail..This will significantly inhibit your ability to grow.. My last bit of unsolicited advice would be to use more conceptual constructs, and as simple as you can imagine them..That being said, thank you for increasing the IQ of RUclips..Peace.
@@lowlize Perhaps you're right..As a retired science professional, I followed most of the content..If the goal is preaching to the choir don't change a thing..IF it's popularizing science, making it more assessable, AND gaining a MUCH wider audience, then changes should be considered..In my humble opinion, of course.
@@Ikbeneengeit With no personal disrespect friend, making a difficult concept EASIER to understand is nearly the opposite of dumbing it down, as you suggest.. Very few would claim that Feynman dumbed down particle physics and their interactions when he created his diagrams..It was a brilliant accomplishment and a more intuitive approach...FLASHING complex equations by the viewer, that NEARLY no one has time to fully look at,, let alone comprehend, certainly isn't adding USEFUL detail to the topic, and arguably introduces more confusion..But I suppose YOU had no such difficulties with any of it..
Thank you for this video, I’ve just subscribed. I’m curious about the production of the videos, as I run a small business and want to create explainer videos with an extremely similar style. Do you do everything yourself or do you have a production company you’d recommend? I currently use Sketchbook Pro and a Wacom One tablet to draw, but I’m really struggling with how to efficiently speed up and slow down the visuals to align with the audio. Maybe this is just something I need to master but yours looks so pro that I’m hoping you’re a brilliant physics person that outsources the production to somebody else that you’d suggest!
@seridian Thank you for the kind words and the support! Unfortunately, I do not outsource my work, I do it all myself. If you have the process of making/recording the drawings down, I highly recommend editing the video in Blender. It's completely free and if all you are doing is speeding up/slowing down and cutting/splicing video segments, it's much faster than other video editing software that I have used in the past. There is a bit of a learning curve to it, but it doesn't take too long to get the hang of. I think I've roughly timed it and a 5-7ish minute video will usually take me like an hour to edit. I'm sure that you could also use something like Adobe Premiere, but I've never used it before, so I can't speak to its efficacy. Sorry I can't recommend anyone, but I hope that helps!
Excellent video man! You just earned a subsciber... I'm guessing you used some python codes to animate the flavor evolution, but can I get some code reference so that I can try this myself? Thanks
@Arun Krishna G Thank you for the kind words! Happy to have you as part of the channel! As for making the animations, I unfortunately seem to have misplaced the code I used but as I recall, I made the visuals for this video in Mathematica, so it may not be what you are wanting anyway... I have since switched to animating in Python (it is generally way faster), so I may still be able to help out! Were you looking more for code as a reference for generally animating in python or for neutrino oscillations specifically? If it is the former, I typically use matplotlib.animation (matplotlib.org/stable/api/animation_api.html) and I actually have a good amount of code at this point that uses this, so I could provide some examples if you want. If you are specifically looking for something which generates these oscillations, I may be able to throw something together for you, but I would need some time if that's okay. Just let me know what all you are looking for!
Been a lot of years since I took P-Chem in college, but I can still follow you. When you say a “new physics” , are you referring to some kind of (Non-Newtonian, Non-Particle) physics? Could it be that we have become “particle chauvinist” in our thinking. And NO. I have no idea what that would look like. Just trying to think outta the box. Or have they thought about that already and I should just continue practicing medicine and stay in my lane.
This is a good question, and I realize now that I wasn't very clear about it in the video. Typically in this sort of context, when we say we need "new physics," we really mean that we should try adding new particle content to the standard model. The question of whether or not we are being "particle chauvanist" (I really like this term btw) is an important one to ask. And the answer is...maybe? The issue with abandoning the particle viewpoint is that the standard model of particle physics is *extremely* successful in almost all cases. So, aside from this issue of the neutrino masses, we find pretty much no definite inconsistencies with the standard model (there are some possible anomalies, but they aren't statistically significant enough to actually make any claims about). To put this into perspective, all of the physics that the standard model predicts all the way back to about 10^-32 seconds after the big bang seems to be pretty much correct. Now, there are other questions which we may ask that might drive us away from a particle viewpoint like explaining dark matter, dark energy, the matter-antimatter asymmetry, inflation, and quantum gravity. It doesn't help that a lot of these questions can be explained using particle physics, especially dark matter. For example, Asaka and Shaposhnikov's neutrino minimal standard model has the potential to explain dark matter, matter-antimatter asymmetry, and neutrino masses by adding 3 flavors of right-handed neutrinos, though it is pretty highly constrained and it isn't the most aesthetically pleasing model. Most likely, it will be the question of quantum gravity that will make us abandon our particle viewpoint, but theories of quantum gravity are extremely difficult to test since most of the results can only really be measured at extremely high energies. So a theory like string theory which is a potential theory of quantum gravity is often criticized for not producing any testable results. TL;DR, particle physics is very convincing, so it is really difficult to just abandon it, though at some level, we probably will have to.
Wait a minute... in 3:37 it is said that in the experiment it is the electron that is observed, and the neutrino is the inferred particle. In 10:55, however, the video seems to imply that the (muon/electron) neutrino is observed. Are these different experiments, or is there more complicated logic here and the second experiment observes muons? And is the statement about the core of the sun heavily theory-based or based on some observational evidence about the particles that the sun is putting out?
Neutrinos interact with the weak force and favour left hand. IMVHO Gravity is exactly this: this type of neutrino friction. Gravity is NOT a force. Wondering: Does the left handedness have anything to do the arrow of time?
Ettore Majorano what a brilliant mind, troubled life and mysterious disappearance. What dimension did you journey to and from on your last voyage? #3767 #PhilosophyOfTimeTravel ...Why is there no movie about him?
@Benjamin Church This just comes from the fact that we use the energy of the neutrino to evolve the neutrino's state in time. In relativity (and using natural units, so c=1), this energy is given by E = sqrt(p^2 + m^2), where E is the energy, p is the spatial momentum and m is the mass. For a neutrino, we expect the mass to be very, very small, so we can assume that p >> m and we can expand the square root in powers of m^2/p^2. To second order, this gives E = p(1 + 1/2*m^2/p^2 + O(m^4/p^4)) and so the energy is approximated by p + m^2/(2p). When we crank through the math and try to find the amplitude for the neutrino to start as e.g. an electron neutrino and end up as a muon neutrino some time later, we find that this depends on Ei - Ej where Ei and Ej are the different energies of the definite-energy states of the neutrino. Since the oscillations should conserve momentum, the leading order term, which is just the momentum, drops out and you are left with 1/(2p)*(mi^2 - mj^2).
@@zapphysics thank you for the explanation. What confuses me is that in the rest frame of the neutrino, the energy states would be Ei = mi and so the phase difference should be Ei - Ej = mi - mj rather than the difference of squares.
Ahhhh, I think I figured it out. If the mass difference is small then mi^2 - mj^2 ~ 2 m (mi - mj) where m is the average mass. For ultrarelativistic neutrinos, p ~ m c gamma and thus (mi^2 - mj^2)/2p ~ (mi - mj)/ gamma agreeing with the earlier result taking into account time dilation
@Thomas Gebhardt this is a phenomenal question, and the answer is actually a bit surprising. It turns out that, in principle, you can work in a system (a basis) where what you call the electron/muon/tau are not energy eigenstates. In fact, you can even choose your basis so that the charged leptons are not energy eigenstates while the neutrinos are! The math all works out the same in terms of e.g. neutrino oscillations, but it can make other things more complicated. Seeing how this works gets a bit tricky, but I will give it a shot. So, say that we combine our three charged leptons (electron, muon, tau) into a three-dimensional vector, which we will call L and the anti-particles are collected into a similar vector called L'. We can do the same thing with the neutrinos and anti-neutrinos by organizing them into 3D vectors which we can call N and N', respectively. Now, we will assign a 3×3 lepton mass matrix, M, and a 3×3 neutrino mass matrix, m. Then, the "energy" terms will look something like L'.M.L and N'.m.N (where the "." represents matrix multiplication). Note that M and m *don't have to be diagonal matrices* and aren't in general. If we were to now "rotate" the vectors L and N by some transformation matrices U and V, respectively so that L->U.L and N->V.N. However, L' and N' turn out to not be independent of L and N and transform as L'->L'.U' and N'->N'.V' where U' and V' are the inverse transformations of U and V, i.e. U.U'=U'.U=V'.V=V.V'=1. Notice that the energy terms now become L'.U'.M.U.L and N'.V'.m.V.N. If this were the end of the story, since U' is the inverse of U, the matrix U can always be chosen so that U'.M.U is diagonal, and similar for V'.m.V, meaning that the transformed L and N are energy eigenstates. The catch comes from the fact that the weak interaction has terms in it like ~L'.N and ~N'.L so that after transformation, these become ~L'.U'.V.N and ~N'.V'.U.L and since U and V are not inverses of each other, these terms *will change under transformation*. This combination of e.g. U'.V is known as the PMNS matrix (the quarks have an analogous matrix known as the CKM matrix which arises for exactly the same reason). If we want to work with flavor eigenstates (i.e. eigenstates of the weak interaction), we would choose U and V so that U'.V=1. However, *this necessarily has to spoil the diagonality of at least one of the two mass matrices*. So we can choose U so that U'.M.U is diagonal but then the requirement that U'.V=1 fixes V so that V'.m.V is not in general diagonal, which directly leads to neutrino oscillations. Note that we get the same answer no matter which basis we choose, it is just that we detect electrons, muons, or taus so it is most convenient to work with the situation where the charged lepton produced in the weak interaction going to be the same ones that we detect, meaning that L must be both a flavor and mass eigenstate, forcing us to choose a basis where N is either a mass eigenstate (and we aren't guaranteed that the flavor of neutrino matches the flavor of charged lepton in the weak interaction) or it is a flavor eigenstate (and the state oscillates as it propagates through spacetime). The end result is the same no matter which picture we choose. One final note: we can easily see that we can diagonalize everything if either M or m are zero. So, if we see neutrino oscillations, we automatically know that neither of the two can be zero! Sorry for the long reply, but this is a very good question and I wanted to answer it adequately!
@@zapphysics Thank you very much for the detailed and comprehensive answer to my question! I have now understood that by a suitable choice of the basis, one can diagonalize only 2 of the 3 relevant matrices (M, m, weak force) if m is not a multiple of the unit matrix. Since the experiments finally do measure charged leptons, it is convenient to describe M and the weak interaction as simple as possible, moving all the clutter to m. In the standard model with massless neutrinos (or even if all neutrinos had the same mass), on the other hand, one can diagonalize everything. The definitions of electron, muon and tau neutrino are chosen exactly so that the weak force preserves flavor and these so defined neutrinos are then automatically energy eigenstates.
This is an amazing display of balancing pop-sci with actual science & math. Thank you for these videos.
This is probably the greatest concise explanation of neutrino oscillations I’ve ever heard. I would love to see another video continuing the story onto sterile neutrinos, neutrino-less double beta decay, and leptogenesis!
Wow, that was an amazing summary of neutrino physics! Also thumbs up for Yoda
If I can shoe-horn in a Star Wars reference, you better believe I will lol!
@@zapphysics Nice
Excited to know I found your gem of a channel before all the rest of the PBS Spacetime subscribers, keep it up!
Damn. You beat me by 3 days
When a dunce such as myself can understand all of this, you need to know you've done an excellent job with your presentation! Perhaps the new physics will be your discovery! 💝
Honestly I really find this as most underrated Physics Channel on RUclips or anywhere... On thr internet...
He posses a quite a grip on language... Explaining everything with Mathematics and Diagram in a way that makes iteasy for anyone to understand and do not get any wrong assumptions regarding the subject/topic which is talked about...
All and all Just keep it up ... bro ....👍👍👍
I learned more in this video than all the Fermilab neutrino videos put together!! SUBSCRIBED!
A elegant argument is that, wheter or not the mass eigenstates change, the fact that there is a variation in neutrino flavor immediately implies they experience time, as per relativity, and only massive particles can experience time.
I think it's shabby, but it's nice sounding.
This video is a capsule form of a semester worth of learning parts of Particle physics. Excellent !!!
I hope that many more people would see that video and many others from this channel
Just stumbled upon your channel and binge watched all of your videos, thank you very much :)
I love the animation at 10:40. For me it was an "aha" moment.
I like the fact that you actually went into some detail. Most channels are wouldn't explain anything from this subject and just say, the standard model is not complete because neutrinos have mass. However, I almost didn't anything in this video because of all the math I had never seen before.
Finally an explanation what "neutrinos must have mass" actually means!!!
Really is an excellent explanation of neutrino and new physics.
The probability is the length squared of that arrow. So the graph at around 10:45 should look different, it should be the square of what's shown. It is a cos function shifted and scalaed to between 0 and 1 given that initial state.
congratulations on 1000 Pi subscribers!
Excellent video
I like this animation and I like this ref. 2:20 :)
OK - I got lost about 2/3 of the film, but still I enjoyed it very much! Great channel.
Regarding neutrino mass it reminds me of another mass that we cannot observe directly - dark matter. Perhaps there is a link there?
@a1nd23 This is a great connection to make! And it has, in fact, been studied. It turns out that the standard model neutrinos (the ones that we *have* observed) are too light to be dark matter. Essentially, in the early universe, they have too much energy to be able to be bound by gravity to form stable galaxies and other large structures. However, it is possible that the sterile neutrinos (the ones we haven't detected) could possibly make up dark matter. In fact, there is a model from Shaposhnikov known as the "neutrino minimal standard model" which extends the known standard model by 3 sterile neutrinos and can explain a whole slew of problems in physics, including dark matter. One big issue with it is that the sterile neutrinos have to be quite massive and have a huge hierarchy (2 neutrinos have to be ~1000 times heavier than the other) in order to actually make the model work out correctly. This is not only aesthetically displeasing, but it also would make them incredibly hard to detect. There also could be some signals from this that we might expect to see, but haven't yet, so that is another issue with it. Either way, it is a possibility, but until anything is detected, we can't know for sure!
Thanks for this!
Personally I enjoyed the show and subscribed..Based on an incomplete data set of a SINGLE video, the following advice could be unjustified but here it is.. Despite your disclaimer, you are STILL getting overly involved in too much detail..This will significantly inhibit your ability to grow.. My last bit of unsolicited advice would be to use more conceptual constructs, and as simple as you can imagine them..That being said, thank you for increasing the IQ of RUclips..Peace.
I think this level of detail coupled with clarity of exposition is what makes this channel special compared to others.
@@lowlize Perhaps you're right..As a retired science professional, I followed most of the content..If the goal is preaching to the choir don't change a thing..IF it's popularizing science, making it more assessable, AND gaining a MUCH wider audience, then changes should be considered..In my humble opinion, of course.
@@lowlize Ps.. Perhaps a well-considered synthesis of BOTH perspectives would be the best solution..
I love the detail, please don't dumb it down, but maybe an overview first and then summary after would help give it structure.
@@Ikbeneengeit With no personal disrespect friend, making a difficult concept EASIER to understand is nearly the opposite of dumbing it down, as you suggest.. Very few would claim that Feynman dumbed down particle physics and their interactions when he created his diagrams..It was a brilliant accomplishment and a more intuitive approach...FLASHING complex equations by the viewer, that NEARLY no one has time to fully look at,, let alone comprehend, certainly isn't adding USEFUL detail to the topic, and arguably introduces more confusion..But I suppose YOU had no such difficulties with any of it..
Thank you for this video, I’ve just subscribed. I’m curious about the production of the videos, as I run a small business and want to create explainer videos with an extremely similar style.
Do you do everything yourself or do you have a production company you’d recommend? I currently use Sketchbook Pro and a Wacom One tablet to draw, but I’m really struggling with how to efficiently speed up and slow down the visuals to align with the audio. Maybe this is just something I need to master but yours looks so pro that I’m hoping you’re a brilliant physics person that outsources the production to somebody else that you’d suggest!
@seridian Thank you for the kind words and the support! Unfortunately, I do not outsource my work, I do it all myself. If you have the process of making/recording the drawings down, I highly recommend editing the video in Blender. It's completely free and if all you are doing is speeding up/slowing down and cutting/splicing video segments, it's much faster than other video editing software that I have used in the past. There is a bit of a learning curve to it, but it doesn't take too long to get the hang of. I think I've roughly timed it and a 5-7ish minute video will usually take me like an hour to edit. I'm sure that you could also use something like Adobe Premiere, but I've never used it before, so I can't speak to its efficacy. Sorry I can't recommend anyone, but I hope that helps!
@@zapphysics thank you so much! Never heard of Blender and definitely prefer free at this stage. Congrats on your diversity of skills!
Excellent video man! You just earned a subsciber... I'm guessing you used some python codes to animate the flavor evolution, but can I get some code reference so that I can try this myself? Thanks
@Arun Krishna G Thank you for the kind words! Happy to have you as part of the channel!
As for making the animations, I unfortunately seem to have misplaced the code I used but as I recall, I made the visuals for this video in Mathematica, so it may not be what you are wanting anyway... I have since switched to animating in Python (it is generally way faster), so I may still be able to help out! Were you looking more for code as a reference for generally animating in python or for neutrino oscillations specifically? If it is the former, I typically use matplotlib.animation (matplotlib.org/stable/api/animation_api.html) and I actually have a good amount of code at this point that uses this, so I could provide some examples if you want. If you are specifically looking for something which generates these oscillations, I may be able to throw something together for you, but I would need some time if that's okay. Just let me know what all you are looking for!
Been a lot of years since I took P-Chem in college, but I can still follow you. When you say a “new physics” , are you referring to some kind of (Non-Newtonian, Non-Particle) physics? Could it be that we have become “particle chauvinist” in our thinking. And NO. I have no idea what that would look like. Just trying to think outta the box. Or have they thought about that already and I should just continue practicing medicine and stay in my lane.
This is a good question, and I realize now that I wasn't very clear about it in the video. Typically in this sort of context, when we say we need "new physics," we really mean that we should try adding new particle content to the standard model.
The question of whether or not we are being "particle chauvanist" (I really like this term btw) is an important one to ask. And the answer is...maybe? The issue with abandoning the particle viewpoint is that the standard model of particle physics is *extremely* successful in almost all cases. So, aside from this issue of the neutrino masses, we find pretty much no definite inconsistencies with the standard model (there are some possible anomalies, but they aren't statistically significant enough to actually make any claims about). To put this into perspective, all of the physics that the standard model predicts all the way back to about 10^-32 seconds after the big bang seems to be pretty much correct.
Now, there are other questions which we may ask that might drive us away from a particle viewpoint like explaining dark matter, dark energy, the matter-antimatter asymmetry, inflation, and quantum gravity. It doesn't help that a lot of these questions can be explained using particle physics, especially dark matter. For example, Asaka and Shaposhnikov's neutrino minimal standard model has the potential to explain dark matter, matter-antimatter asymmetry, and neutrino masses by adding 3 flavors of right-handed neutrinos, though it is pretty highly constrained and it isn't the most aesthetically pleasing model.
Most likely, it will be the question of quantum gravity that will make us abandon our particle viewpoint, but theories of quantum gravity are extremely difficult to test since most of the results can only really be measured at extremely high energies. So a theory like string theory which is a potential theory of quantum gravity is often criticized for not producing any testable results.
TL;DR, particle physics is very convincing, so it is really difficult to just abandon it, though at some level, we probably will have to.
Wait a minute... in 3:37 it is said that in the experiment it is the electron that is observed, and the neutrino is the inferred particle. In 10:55, however, the video seems to imply that the (muon/electron) neutrino is observed. Are these different experiments, or is there more complicated logic here and the second experiment observes muons?
And is the statement about the core of the sun heavily theory-based or based on some observational evidence about the particles that the sun is putting out?
That quickly turned into pure mathematics
Neutrinos interact with the weak force and favour left hand.
IMVHO Gravity is exactly this: this type of neutrino friction. Gravity is NOT a force.
Wondering: Does the left handedness have anything to do the arrow of time?
Sir, I am a big fan and I study physics myself but this is kind of next level for public and I recommend you to explain smaller concepts first
But photons can change particle types too...
Ettore Majorano what a brilliant mind, troubled life and mysterious disappearance. What dimension did you journey to and from on your last voyage? #3767 #PhilosophyOfTimeTravel ...Why is there no movie about him?
Who is this?
@@avnishbadoni1393 He is the one that Vu physics mentions and named after him the majorano neutrino etc.
m^2 t does not seem to have correct units! L/2E ?
can you explain why the mass is squared in the frequency at 8:30?
@Benjamin Church This just comes from the fact that we use the energy of the neutrino to evolve the neutrino's state in time. In relativity (and using natural units, so c=1), this energy is given by E = sqrt(p^2 + m^2), where E is the energy, p is the spatial momentum and m is the mass. For a neutrino, we expect the mass to be very, very small, so we can assume that p >> m and we can expand the square root in powers of m^2/p^2. To second order, this gives E = p(1 + 1/2*m^2/p^2 + O(m^4/p^4)) and so the energy is approximated by p + m^2/(2p). When we crank through the math and try to find the amplitude for the neutrino to start as e.g. an electron neutrino and end up as a muon neutrino some time later, we find that this depends on Ei - Ej where Ei and Ej are the different energies of the definite-energy states of the neutrino. Since the oscillations should conserve momentum, the leading order term, which is just the momentum, drops out and you are left with 1/(2p)*(mi^2 - mj^2).
@@zapphysics thank you for the explanation. What confuses me is that in the rest frame of the neutrino, the energy states would be Ei = mi and so the phase difference should be Ei - Ej = mi - mj rather than the difference of squares.
Ahhhh, I think I figured it out. If the mass difference is small then mi^2 - mj^2 ~ 2 m (mi - mj) where m is the average mass. For ultrarelativistic neutrinos, p ~ m c gamma and thus (mi^2 - mj^2)/2p ~ (mi - mj)/ gamma agreeing with the earlier result taking into account time dilation
I wonder why there are no electron/myon oscillations, i.e., why these lepton flavors coincide with their energy eigenstates.
@Thomas Gebhardt this is a phenomenal question, and the answer is actually a bit surprising. It turns out that, in principle, you can work in a system (a basis) where what you call the electron/muon/tau are not energy eigenstates. In fact, you can even choose your basis so that the charged leptons are not energy eigenstates while the neutrinos are! The math all works out the same in terms of e.g. neutrino oscillations, but it can make other things more complicated.
Seeing how this works gets a bit tricky, but I will give it a shot. So, say that we combine our three charged leptons (electron, muon, tau) into a three-dimensional vector, which we will call L and the anti-particles are collected into a similar vector called L'. We can do the same thing with the neutrinos and anti-neutrinos by organizing them into 3D vectors which we can call N and N', respectively. Now, we will assign a 3×3 lepton mass matrix, M, and a 3×3 neutrino mass matrix, m. Then, the "energy" terms will look something like L'.M.L and N'.m.N (where the "." represents matrix multiplication). Note that M and m *don't have to be diagonal matrices* and aren't in general. If we were to now "rotate" the vectors L and N by some transformation matrices U and V, respectively so that L->U.L and N->V.N. However, L' and N' turn out to not be independent of L and N and transform as L'->L'.U' and N'->N'.V' where U' and V' are the inverse transformations of U and V, i.e. U.U'=U'.U=V'.V=V.V'=1. Notice that the energy terms now become L'.U'.M.U.L and N'.V'.m.V.N. If this were the end of the story, since U' is the inverse of U, the matrix U can always be chosen so that U'.M.U is diagonal, and similar for V'.m.V, meaning that the transformed L and N are energy eigenstates. The catch comes from the fact that the weak interaction has terms in it like ~L'.N and ~N'.L so that after transformation, these become ~L'.U'.V.N and ~N'.V'.U.L and since U and V are not inverses of each other, these terms *will change under transformation*. This combination of e.g. U'.V is known as the PMNS matrix (the quarks have an analogous matrix known as the CKM matrix which arises for exactly the same reason). If we want to work with flavor eigenstates (i.e. eigenstates of the weak interaction), we would choose U and V so that U'.V=1. However, *this necessarily has to spoil the diagonality of at least one of the two mass matrices*. So we can choose U so that U'.M.U is diagonal but then the requirement that U'.V=1 fixes V so that V'.m.V is not in general diagonal, which directly leads to neutrino oscillations. Note that we get the same answer no matter which basis we choose, it is just that we detect electrons, muons, or taus so it is most convenient to work with the situation where the charged lepton produced in the weak interaction going to be the same ones that we detect, meaning that L must be both a flavor and mass eigenstate, forcing us to choose a basis where N is either a mass eigenstate (and we aren't guaranteed that the flavor of neutrino matches the flavor of charged lepton in the weak interaction) or it is a flavor eigenstate (and the state oscillates as it propagates through spacetime). The end result is the same no matter which picture we choose.
One final note: we can easily see that we can diagonalize everything if either M or m are zero. So, if we see neutrino oscillations, we automatically know that neither of the two can be zero!
Sorry for the long reply, but this is a very good question and I wanted to answer it adequately!
@@zapphysics Thank you very much for the detailed and comprehensive answer to my question!
I have now understood that by a suitable choice of the basis, one can diagonalize only 2 of the 3 relevant matrices (M, m, weak force) if m is not a multiple of the unit matrix. Since the experiments finally do measure charged leptons, it is convenient to describe M and the weak interaction as simple as possible, moving all the clutter to m.
In the standard model with massless neutrinos (or even if all neutrinos had the same mass), on the other hand, one can diagonalize everything. The definitions of electron, muon and tau neutrino are chosen exactly so that the weak force preserves flavor and these so defined neutrinos are then automatically energy eigenstates.