The idea is that every event (eg. a particle collision) sends waves forwards and backwards in time, eg. if * is an event and </> are waves moving backwards/forwards in time:
<- past future ->
<<<<<<<<<*>>>>>>>>>
The waves from multiple events can overlap and interfere, eg.
<<<<<<<<<<<<*>>>>>
<<<*>>>>>>>>>>>>>>
The interference depends on the phase of the waves, but let's simplify and say that similar arrows are destructive (represented as a space) and opposite arrows are constructive (represented as a dash). In which case, the interference pattern of the example above would be:
*--------*
It looks like there is something which is created at the first event, travels through time to the second event, and is then destroyed. That "thing" is what we'd call a particle. This idea is called "transactional" because it treats the existence of a particle as not just depending on the event which creates it, but also on the event which eventually destroys it, and the interference of these "waves through time" is like a 'negotiation' between the two events.
It'd be interesting to know how they deal with the phase requirements--normally you have boundary conditions which set up these sorts of standing waves but it sounds like these "events" would not have that and thus would have to conspire their phases so that you wouldn't get something more like:
---* *---
Probably it's something like, "each event must exist on at least two different fields; if you look at the other (~) field too then this event looks like
---*~~~~~~~~*---
so that, for example, this electron clearly collided with a high-energy photon to become a muon for a time and then relaxed back to being an electron, emitting the photon back."
I'm with Feynman. One electron sounds like nonsense. (Where exactly does the "looping around" happen, where it changes directions? Outside of time?) However, the idea of positrons as electrons moving backward in time is cool, and does seem like a convenient sort of an explanation for their creation and annihilation.
The idea is that any time an electron-positron pair is created or destroyed is a change in direction.
Say a pair is created in space, they move around for five seconds, and annihilate each other. You could also explain that as a single electron in a loop – electron moves forwards in time for five seconds, changes direction (becomes a positron), moves back in time for five seconds, changes direction, the cycle repeats.
You can extend that to a set of two of electrons/positron pairs. Assuming each electron is created and destroyed with different positrons, you can again explain the system in terms of a single electron that changes direction four times. Add a few more in and you have a whole universe with one electron.
(That assumption may be a big one, which is where things fall down, but it's not nonsense. Also, bear in mind that the while idea of "moving" back in time is a helpful analogy, it's not very precise and shouldn't be taken too literally)
Going a little further with the thought (and I am definitely not a physicist), what if that electron "loop" (or even the "single electron") is itself moving within the four-dimensional space. Could that explain the particle wave function? Meaning a particle's location is expressed as a probability because its location really is random-ish (i.e. the loop is moving around) until it's "locked in" within our subjective "arrow of time" frame of reference by a particle interacting with it.
P.S. I acknowledge that I have absolutely no real-world basis for this idea, just a thought.
Of course, yes. Thank you. For some reason my mental picture was limited to creation and annihilation events being the exception, with most particles existing 'throughout time', but that's what's nonsense.
In fact, if you consider that a lot of particles would have presumably been created in the big bang, at which point they would conveniently be in the same location in our other three dimensions as well, it's certainly conceivable that you could have a single particle pinballing back and forth as you (and I guess Wheeler) describe.
Of course, still far from a certainty, but a neat idea.
What is this time thing you speak of? What is this "inside" of time that you imply the existence of?
An extrapolation from the one-electron-universe hypothesis is that the entire universe exists within a single point - this is why constants are constant across the universe, why matter is the same everywhere, why particles share properties, how tunnelling works, etc. - we just perceive it as being uncompacted from our reference frame, whatever or wherever that may be. Elements of some SUSY theories touch on this too - that there is only one string, that the universe is a projection from higher dimensions "curled up" in a planck-scale point....
Frankly, we can't know, and I'm not sure that we ever really can, due to the limitations of our faculties, and the rooting of our "reality" in how this particular collection of molecular machinery interacts with other physicality around it.
It's a (mathematically) multidimensional thing; if things can "move" there are clearly dimensions those things can be spread across.
There are lots of things we can't know, but it's like toothpaste; if we posit certain things as fixed, if our observations mean anything, something else has to vary.
> There are lots of things we can't know, but it's like toothpaste; if we posit certain things as fixed, if our observations mean anything, something else has to vary.
...how is that like toothpaste?
I don't disagree with what you're saying, but... toothpaste?
This makes me wonder: how would a change in direction in/on of the other axes (or manifolds?) on the "remaining" proposed 7 dimensions manifest themselves? Does stuff actually move on those axes or over/through other manifolds?
Thanks - you have now rendered me incapable of doing anything productive for the rest of the work day as I stare out the window contemplating that concept.
I like the idea - why should everything be a different thing? I think that perhaps all things are actually different aspects one thing. It makes the universe much simpler.
I love the direct quote: "I did not take the idea that all the electrons were the same one from [Wheeler] as seriously as I took the observation that positrons could simply be represented as electrons going from the future to the past in a back section of their world lines. That, I stole!"
I can absolutely hear it in Feynman's rapid, wry, dramatic voice.
You can read his whole Nobel speech online and it's well worth a read. My favorite part is:
"There was a gentleman, newly arrived from Europe (Herbert Jehle) who came and sat next to me. Europeans are much more serious than we are in America because they think that a good place to discuss intellectual matters is a beer party. So, he sat by me and asked, "what are you doing?" and I said, "I'm drinking beer!" Then I realized that he wanted to know what work I was doing..."
One-electron seems very problematic in light of the observed antimatter asymmetry. Why/how could the electron travel forwards in time more often than backwards?
Well, assume that time loops back onto itself. Then the electron goes forward most of the time and you see multiple copies. ( This is of course ridiculous, everybody knows that all electrons are just pointers to one const particle structure. )
Obtaining multiple electrons by forcing a single one to bounce back and forth all the way across the time line a zillion times seems like a major design smell. That would be the worst hack ever, or at least the worst one before Facebook devs hacked Dalvik just to get their app to run on Android: http://jaxenter.com/facebooks-completely-insane-dalvik-hack-... ;)
Perhaps our universe is one created by a junior, or an intern, and this "creative" workaround got mocked on TheDailyWTF somewhere.
Well, it's immutable. So why would it be a problem with bouncing it around?
It's like a symbol in Ruby -- all instances of, say, :electron point to the same point in memory, they're just used all over the place. The sharing isn't problematic because you can't change :electron.
Why would you want to use 'electron' instead where every time you use it, you put another instance of it into memory?
It's only a "major design smell" to "bounce back and forth" if it's computationally expensive. This theory makes it sound like electrons aren't firmly rooted in space and time to begin with.
Maybe it's computationally cheaper to define something as existing in all places and all times.
Well I imagine it isn't optimized for readability :)
If the point is to have a jillion of electrons, each at the right time and place (for the observer), I assume a rather complicated mechanism must have been put in place to "tie the knot" just right - I mean to "navigate" the time-travelling electron just so it never fails to appear wherever, whenever it's expected.
It would be terrible to design it that way, but I think it's fine as a compiler optimization. Most of that datastructure is constant anyway, think of the memory savings!
In that case there'd be no need to convert back-and-forth between electrons/positrons or forward/reverse time at all. The only requirement would be that the observed Universe is a fixed-point of this loop.
Well, perhaps it's only in our time frame. For instance, let's imagine that the electron goes forward from 0 to 5, then bounces back to 1, then to 5 again, and back to 2 etc.
If we were at 1.5, we would see it moving forward TWICE (0=>5 and 1=>5), but only returning ONCE (1<=5; we haven't gotten to observe 2<=5 yet).
Whether it's plausible I have no clue :) it just shows that in principle you can have local asymmetry while everything still balances out at the end.
That doesn't work because you've left the loop open. If the positron just goes back to 2 and stops, from our point of view it has just appeared from nowhere, which isn't physical.
In order to close the loop you have to go back to 0 and meet the original electron. Once you've done that you lose the asymmetry again.
Well, I won't pretend that I know how it's supposed to work, but even if the loop was closed, the electron (and the loop itself) still must have come to existence out of nowhere somehow, so it doesn't remove this ontological problem :)
I'm not really talking about an ontological problem though, I'm saying that an electron appearing on its own is a physical impossibility. Electrons must appear or disappear alongside a positron partner, which means that the idea of "local asymmetry" can't solve the symmetry problem.
> Any given moment in time is represented by a slice across spacetime
Is this actually true when we take into account special relativity? I've always struggled with this.
I was taught that this sort of thing is not nearly so simple. For instance, if you fix a point in spacetime and have a timelike vector, representing the motion of an observer, then the set of events which that observer will perceive as simultaneous to the fixed point all lie in the plane orthogonal to that timelike vector. Even more bizarrely, because this is a non-Euclidean notion of orthogonal, this orthogonal complement rotates towards the vector as it deviates from the centerline of the future line cone at that point.
Given all this, I don't see how one can just say that a slice across spacetime somehow represents a particular moment in time. I thought the whole point of relativity is that there are no absolute time slices in spacetime.
Yes, it's still true when we take special relativity into account. One good perspective on special relativity is that it sort of says, "everybody's right." That is, you abstract your "present" moment to distant positions as a 3d plane in the 4d space; someone else at the same position with a different velocity will have their present too, and those will not be 100% commensurate, but that's OK -- both of you have valid coordinate systems for any calculations you want to accomplish.
The family of all the "presents" of a point in spacetime is the set of all points which are spacelike-separated from that point -- i.e. if that point is at the origin of some coordinate system, the complete "relativistic present" is all of those points such that
c² t² − x² − y² − z² < 0
Every observer passing through a point agrees upon these points exactly; the only difference is that my t=0 will not correspond to t' = 0 under a Lorentz transform, so that my "simultaneous" at space-separated points is not someone else's "simultaneous" at those points.
(Light bubble picture: imagine that the light which shines upon an event in spacetime expands outwards with speed c, forming an expanding bubble of light. Consider two events. They are both "simultaneous" in the sense that there is an observer who thinks that they are simultaneous, if their light bubbles start out topologically disconnected and overlap eventually. They are both "time-ordered" in the other case, if one bubble is inside the other. If two things are objectively time-ordered then they are not objectively space-separated, because the points in the larger bubble correspond to valid inertial trajectories of a spaceship going less than the speed of light -- there are some spaceships which visited both events inertially. Similarly if two things are objectively space-separated then they are not objectively time-ordered; consider someone on the intersection of the two light bubbles seeing both events happen "right now"; there is always a velocity vector such that they will trace the distances back to the origins of the events as equal -- and hence that observer thinks that both happened simultaneously at their different locations.)
Not being a physicist (not even close), I find it amazing how within just ~100 years, "our" model of the universe has changed so drastically.
I once read that sometime in the late 1800s, people assumed that they had basically understood how the universe works, and that all that was left was to fill in some of the blanks.
And now look what an increasingly strange (and wonderful!) place we find ourselves in.
Well I suppose there is a way to disprove this: create an electron-positron pair, then annihilate the pair. This way we would get a closed loop, which would obviously not be a part of a giant knot spanning through the universe.
It's not "obvious" though: there is no proof that your electron didn't (via tunneling) exit whatever trap you put it in only to get replaced by some other electron from the environment which jumped in a short moment after (also via tunneling). The best you could say is "it's highly probable that at least at one time it was at least a two-electron universe."
>"the eventual creation and annihilation of pairs that may occur now and then is no creation or annihilation, but only a change of direction of moving particles, from past to future, or from future to past."
Forgive my ignorance, but how does the fact that annihilation "produces" energy out of the two particles fit into this?
Since a photon is its own antiparticle, you might imagine that a photon and an electron "bounce" off of each other in time, and both reverse their directions. From our point of view it looks like two photons created and two electrons annihilated, or vice versa.
Mostly a guess: the energy produced by annihilation is the energy required to reverse the direction of the electron. Similarly, the energy require to create a pair is the same quantity.
Every time some limit is reached in a physics equation, dilation, contraction or even moving in time is taken as a measure to rescue. It all sounds like a convenient method for explaining something we can not explain. It all comes from the fact that time is defined through speed and speed depends on space, thus time cannot describe dimensions not dependant of space.
I believe you are referring to velocity = distance/time. This is the non-relativistic view of velocity. It makes more sense when you define what distance is.
When you push something with your finger and move it, you do not change it's velocity. You change it's acceleration. It's a small but important distinction. Under relativity, the distance between two points in spacetime is defined by the minkowski metric:
d = sqrt( x^2 + y^2 + z^2 - (ct)^2 )
Notice time is negative/imaginary. The tl;dr of this is that an object at rest is still has a velocity along the time plane at the speed c, and any non rest velocity is relative to that (which affects d), which is where length contraction comes from. You're right in you can't define one without the other, so it's redefined around the constant c and the minkowski metric. It's why they call it spacetime.
In what sense is time defined by speed? Surely it's the other way around?
Currently the second is defined by a number of oscillations of a known wavelength of radiation, which explicitly avoids any dependence on space / measurement of spatial speed.
Yes, but the standard second doesn't depend on wavelength (maybe I should have said "known energy"). You take a Caesium-133 atom, look at it's emission spectrum, isolate the ground state radiation and time a number of oscillations. Of course, the radiation has a specific wavelength/energy, but the calculation doesn't depend on it.
Once you know about time, you can start talking about distance – a meter is defined as the distance light travels in a given time. Given that the meter is defined this way, defining time in terms of speed would be a bit circular.
Ok, the time dilation makes actually sense now. So like chemical reactions depend on temperature and pressure, physical reactions (if they can be called this way) are dependant on gravity and velocity (dependance on the last I can not quite get still). By physical reactions I mean the transitions between the ground states that time is defined by.
Sure, local velocity at any rate. If someone goes past you in a rocket, you'll notice that their Caesium radiation is oscillating slower than yours, so their second will appear to take longer than yours.
I received a telephone call one day at the graduate college at Princeton from Professor Wheeler, in which he said, "Feynman, I know why all electrons have the same charge and the same mass" "Why?" "Because, they are all the same electron!"
The idea is that every event (eg. a particle collision) sends waves forwards and backwards in time, eg. if * is an event and </> are waves moving backwards/forwards in time:
The waves from multiple events can overlap and interfere, eg. The interference depends on the phase of the waves, but let's simplify and say that similar arrows are destructive (represented as a space) and opposite arrows are constructive (represented as a dash). In which case, the interference pattern of the example above would be: It looks like there is something which is created at the first event, travels through time to the second event, and is then destroyed. That "thing" is what we'd call a particle. This idea is called "transactional" because it treats the existence of a particle as not just depending on the event which creates it, but also on the event which eventually destroys it, and the interference of these "waves through time" is like a 'negotiation' between the two events.