Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Can time have a direction

  1. Apr 28, 2013 #1
    I recently had this idea (which perhaps is not a new one, but I have never heard of it), that on the scale of the entire universe, time expands in all directions at once. I believe it can be proven or disproven with more knowledge, so I am posting it here.

    If we look around us, we generally understand time as 'direction of changes' and everything we see moves in this same direction at the same 'pace'. An hour for me is an hour for you etc. Thanks to Mr Einstein we know this is not a precise description, but from what I understand, if we ceased all movement (relatively of course), time would 'flow' the same for all non-moving objects.

    Now, I was thinking, maybe that is wrong. At least, on the scale of all universe. For some reason I can't attach any image, so if you are interested, look at this crudely drawn picture:
    It is simplified, but the circle represents the universe as it is 'now'. It could be slightly better to imagine it as a 'bubble', but much harder to draw. Please have look and place our Sun and us at point 'A'.

    The range between 'D' and 'J' is the observable universe - what's on the other side 'didn't happen' - 'isn't happening' for us. The strange thing is, though, that if we were a piece of matter at point C, we could see, and even be influenced by, points D and E which is 'invisible' for us.

    This idea also fits nicely with the expansion of universe and even with the dark energy.

    If it is right, then a galaxy 6 billion l. years away is not only 6 billion years younger. Time for this galaxy, relatively as we see it, would actually pass slower because of the distance. I am sure that all of you experts in cosmology will swiftly disprove me and my idea will degenerate into yet another fruitless mental exercise, but still, I thought it may be interesting. And I believe we have so little data about objects billions years away that proving how does the time pass there may be challenging.

    Regards and have fun :) Sorry if my English wasn't perfect.
    Last edited: Apr 28, 2013
  2. jcsd
  3. Apr 28, 2013 #2


    User Avatar
    2017 Award

    Staff: Mentor

    To verify or disprove anything, you first have to define what "time expands in all directions at once" is supposed to mean. Does it give specific predictions for observations? Are they different from standard cosmology? If the first question can be answered with "no", it is pointless to discuss it. If the second answer is "no", it could be (but does not have to be) a nice model to consider.
  4. Apr 29, 2013 #3


    User Avatar
    Science Advisor

    The dimension of time is orthogonal to the dimensions of space. Yes, time has a direction, but that direction is not up or down, left or right, forward or backward. It's a fourth direction of motion: future/past. And it does have a direction given by changes in entropy. That is to say, the past is different from the future because entropy was lower in the past.
  5. Apr 30, 2013 #4


    User Avatar
    Gold Member

    Perhaps you have something in your mind, what people call cosmic time?
    At any specific instant of cosmic time the galaxies would measure the same scale factor, which represents the expansion of the universe.
  6. Apr 30, 2013 #5
    Sorry for the delay. I like to think on my answers. I also have to use proxy to use this site, and it is discouraging. I swear it is not because I did anything to upset the administration of this forum. I have different IP with each connection, but it seems my entire provider is banned here (???).

    By "direction" of time I meant, of course, 'from the past into future', but I also meant 'at a certain relative speed'. Perhaps a 'vector' (not 'scalar') would be a better word. I agree with what Chalnoth wrote, but nothing I posted above is in opposition to his description of time. (I understand this is important to define, that is why the topic title is formed as a question).

    I see time as a vector, and I dare assign 'direction' to it, because it is not as universal as we thought before STR. If person A looks at B and sees his watch moving slower, it is as if person B is heading in a different time direction, moving along different vector. 'A' can only see that component of vector b which is parallel to his own movement. But for 'B', of course, time is not slowed down.

    By time moving in 'all directions' I meant that on the scale of a whole universe, it can not only slow down, relatively, but even work 'backwords'--'towards the past'. And we are not capable of seeing it (obviously). But person B could just catch a glimpse of what is hidden from A because of the difference between his and A's direction.

    Isn't it called a time frame, or am I mixing ideas? I am sorry if I am forging too many terms for the same thing :(

    Mfb, as to your specific questions. Quite obviously, I do not want to try and prove something that is against established facts (here goes the second question). I hope that when you point a telescope at a galaxy 1 billion years away, you don't only see it as it was long time ago, but also you see time moving slower there (if you could see such details). Otherwise, my idea makes no sense, because it makes it sort of inherent property of space, that each point moves along slightly different direction in time.

    Now, does it give any specific predictions?

    The idea means that it doesn't even matter if two points are moving at different relative speeds--they are in different points of space and that is enough for them to have different vectors of time. At least that's what would happen if there was no other force at work (like gravity).

    The typical understanding is that 'space expands' and the galaxies 1 billion years away are speeding away so fast that when we look at them, we see time slow down.

    Assuming the time vectors are not 'chaotic' but, throught the universe, they align into a concentric shape, it means that the farther things are, the more time needs to slow down (from our perspective). And I would see it as a reason, not the result, of expanding space. It expands, because time (on the scale of whole universe) is not moving in one 'direction'. So the space compensates by stretching.

    All things said, I think this is not really anything new, but rather a nice new perspective on what is going on with the universe. you know, mfb, ideas don't have to immediately 'give specific predictions for observations'. When Copernicus said we move around the Sun, not Sun around us, it was only a shift in perspective. But it was huge...

    For example, it might be easier to explain the existance of forces invisible to us, like dark energy. Space, especially distant parts of the universe, could be under influence of objects which we cannot percieve, because they are 'frozen' from *our* perspective. The same with the edge of observable universe. I like to compare it to a complex picture on a thin paper, seen from an angle that allows us to only see the edge. But as we jump several bilion l. years to another galaxy, this edge turns out to be several billion years farther, and and we suddenly are influenced by forces which we couldn't previously percieve.

    And we wouldn't need to look for an answer to 'why is the space expanding'. We would only have to find out why time exists :)

    @timmdeeg: I saw your reply only now. I will definitely look into the info you provided, thank you.
  7. Apr 30, 2013 #6


    User Avatar
    2017 Award

    Staff: Mentor

    Due to the expansion of space, everything is redshifted (seems to be slowed down). This does not depend directly on the age, however, it is a property of the expansion speed. You can even reproduce a similar effect on earth with the Doppler shift.
  8. Apr 30, 2013 #7


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    If you PM me your IP address, I can look into it. First make sure that you're currently unable to access Physics Forums without a proxy. Then go to the following site to find out what will be the source address of IP packets that are sent to our server when you don't use a proxy:


    Copy the IP address at the top of the page and either PM it to me, or hit the report button next to this post, paste the IP address into the box that appears, and submit the report. This sends the report to a place where all the moderators (and no one else) can see it. It quotes this post automatically (if you specifically use the report button next to my post), so you don't have to type an explanation.
  9. May 7, 2013 #8
    @ Fredrik: thanks, it seems the problem has been taken care of.


    redshift and 'seems to be slowed' are only manifestations of a deeper problem, which is what I am trying to talk about, that time 'points' in different directions depending on where in space you are. During 100 years that passed for me, my alien friend in a galaxy far, far away seems to have aged only 50. The same is the other way around. You could say we are 'pointed' in different directions, different (shifted) futures.

    Now, have you ever tried to imagine it the other way around: that it is the reason why galaxies drift apart?

    @ timmdeeg: thank you for your contribution, the Wikipedia entry was too short really but I googled some details, It is generally the same concept (I also assumed 'homogeneousness' of the universe for the sake of simplicity). But I see very little about that in the Internet, except using it as a way to point to the origin of Big Bang. In my opinion it can lead to more interesting ideas.
  10. May 7, 2013 #9


    User Avatar
    Gold Member

    Whitefire, would you agree that the direction of time is given by the sequence of cause and effect?
    If yes, then it seems a bit dubious to consider time as both, cause (you said reason) and direction.

    Perhaps I misunderstood. Are you saying time is the cause that the galaxies are receding?
  11. May 7, 2013 #10
    lets take your analogy of times varying degrees of direction. This would imply that time has a preferred direction depending on the observer.

    We measure the distance from one side of the sun, distance = x. We then measure that same distance from the other side of the sun. If time had a vector direction other than forwards and backwards. We should end up with a different value, based on current mathematics of time. If time had the vector like properties in your descriptive. We would need to factor time as a vector each and every instance we take a measurement. OK lets take measurements of the CMB. If we look in different directions according to your analogy, we should arrive at different time values for the time that the CMB formed.
    Take this a step forward. Lets compare two readings of a distant object. Take one reading from Earth. the other reading from a space probe such as Voyager 1. If time were a multiple angle vector we would have two different values depending on the angles between the two measuring devices.

    None of this occurs, why ? simply put time is not radiating outward from some central point. Its only two valid vectors is the vector representing forward and the vector representing its direct opposite.
  12. May 7, 2013 #11


    User Avatar
    Science Advisor

    What you have stated is a highly non-rigorous and ambiguous statement. If we are dealing with a FLRW universe ##(M,g_{ab})## then there exists a one-parameter family of space-like hypersurfaces ##\Sigma_{t}## that foliate ##M## and the global isotropy implies that the 4-velocity field of the family of isotropic observers ##u^{a}## is hypersurface orthogonal to ##\Sigma_{t}## meaning that at any given ##t_{0}## and any ##p\in \Sigma_{t_{0}}##, ##u^{a}## is orthogonal to all ##s^{a}\in T_{p}\Sigma_{t_{0}}##. You cannot just non-nonchalantly throw around the word "time" and "direction of time" in general relativity.
  13. May 7, 2013 #12


    User Avatar
    Science Advisor

    If time is to point in any direction at all, it points in the "time" direction, which is orthogonal to the three dimensions of space. It simply doesn't make sense to talk about time as pointing in a spatial dimension.
  14. May 8, 2013 #13
    More like 'the way time works on the scale of the whole universe is the cause that galaxies are receding'. Locally, meaning more or less within a single galaxy, it would be like trying to see the curvature of earth by climbing a tree, with stars' gravity acting like mountains to obscure the view.

    @Mordred: It occurs, the problem is probably with my poor choice of words. Just not really on the scale you write about (see paragraph above). Let's say you measure the distance from Milky Way to A--it is 5x10^25 meters away. Now to point B--it is 10x10^25
    meters away, exactly twice the distance--looking from Milky Way. If (hypothetically) you
    find the way to make a measurement while being in galaxy A *also now*, you will not get 5x10^25 to galaxy B. Because in the time it took the light to travel from B to A, and then from A to Milky Way... etc.

    It's like saying you always move from behind and go forward. True, but not exactly about directions. When time passes, there's a change. You can say you move along it. If you add relativity and movement, what you get is very much like a direction, or vector, in 3D. If there is a better term for it though, just say it.



    So maybe this is a problem. I made a visualization, flattened time and space into a 2 dimentional picture, and then I started to rely on it too much. I promise, when I learn a "less non-rigorous, less ambiguous and less non-nonchalant" way to say it, I will use it, unless it is so numerical it will obscure any real understanding. By all means, teach me.

    But I don't know it yet, so I will stick to it for a while. From the past, into the future, at the speed of 1 second per second, strange as it sounds. That is what I mean by 'direction', and I am growing sure that it is something unique for each point in space.

    If you prefer, and if you still have the patience, I can present what I drew on the pictures without using the word 'direction'.

    1) If you look at a point halfway to the edge of the visible universe, 1 second for us is only 1/2 second there. Just this fact for a moment, forget about galaxies' movements and redshifts.

    2) From 1) I conclude that a perfect 1 second per second is relatively ("if you look...") possible for only one the point in space. The observer.

    3) Let's have 2 observers. They are in different points in space, so for each 10 seconds they measure for themselves, they only measure 9 seconds for the other one.

    4) As the seconds pass, space between the observers has to stretch to provide the necessary relative movement. If the distance between them was the same at the beginning and the end of measurement, it would mean that time actually, non-relativistically passes at various speed at 2 points. It doesn't. So we're getting redshifts and movements.

    5) I still think that presenting it as a picture with arrows and directions was a nice way :) For example, it is easy to see that the curvature from the picture can be nicely finished to form a circle.

    @ WannabeNewton: ...translating...


    wow, seems like a very fancy way of really saying NOW. I wouldn't be able to put it this way, which is a pity. Thinking about relativity really messes with what one can understand by 'now'. So precise is good. But what you write is what we really intuitively understand by 'now'. A caveman's wife would wait for him, tending to a little caveling, and think: 'what is my husband doing NOW'? And by 'now' she would just imagine the same time for him as it was for her. She wouldn't think that she can't really see him *immediately* even if he appeared. She just imagined this 'now'.

    I translated 'orthogonal' (sorry for my language problems, I thought 'paralell' is the only term in English for that). Are you saying that for any observer in this imaginary 'now', time moves at the same rate? If I am right and that is what you are saying, than I agree. If you look at the picture in this post you will also see that for both A and B 10 seconds pass. I understand that. But A's 10 seconds aren't necessarily B's 10 seconds. What if your family of isotropic observers started to observe one another, not themselves? It is as if they move the same distance, but at different angles. In different directions...? Curse, that word again. Help. Replace it :) Different... what.
  15. May 8, 2013 #14
    There is an old expression " In order to change the rules, you must first understand the rules"

    For one cosmology does measure the curvature of the cosmos. Its extremely close to flat, so close, that for all practicality it is flat.

    Secondly we have what is called comoving distances as opposed to proper distances. What your describing doesn't even work in GR terms.

    The current FLRW metric accounts for GR affects on large scales. (FLRW metric assumes homogeneous and isotropy on large scales.)

    In order for your time direction analogy to work it would need to work at any scale even at the quantum level. Defining time is that critical.

    here is a link on light cones. Might be a starting point to understand WannabeNewtons post


    this article is technical but it covers the mathematics of light cones.


    by the way this is a cool link on light cones in different applications

    that should cover light cones now lets look at time.

    http://philosophyfaculty.ucsd.edu/faculty/ccallender/index_files/time%20in%20physics.doc [Broken]

    now whats the difference between a scalar quantity and a vector quantity?


    time is a scalar not a vector property.

    As stated at the beginning one needs to understand the rules before you change them
    Last edited by a moderator: May 6, 2017
  16. May 9, 2013 #15


    User Avatar
    Science Advisor
    Gold Member

    Well, we already have a well defined direction for entropy, you cannot unbreak an egg, and you cannot unbreak the time line of the universe. Mathematically, we can run the universe backwards, but, our ability to predict the future is very limited and model dependent. I view that as insisting processes in the universe are irreversible and have an indeterminate future. To suggest otherwise suggests the universe is illogical [and filled with paradoxes] - and also suggests the scientific method is invalid. I fail to see how that would be meaningful.
    Last edited: May 9, 2013
  17. May 9, 2013 #16
    Hey, no one's trying to unbreak the eggs :) For every egg there is, time moves 100% forward. Sounds like a nice slogan. Mordred, I looked up some of your links, only some so far, but I will get to it, promise.

    Let's say we have a past light cone of a galaxy. Well, let's simplify it to a point. Is it possible for 1 galaxy to be in this light cone, while another (galaxy) is not (in the same 'hyperspace of present')?
  18. May 9, 2013 #17


    User Avatar
    Gold Member

    Well said.
    In my opinion this together with Chalnot's short and nevertheless sufficiently specific comment contains all neccesary for Whitefire to reconsider his point of view.
  19. May 9, 2013 #18


    User Avatar
    Gold Member

    Another approach - and still in the context of 'time is orthogonal to space' - would be to neglect the inhomogeneities and instead stay with the perfect fluid model, which expands adiabatically with constant entropy. With k = +1 cosmic time would run backwards in the recollapsing epoch.
  20. May 9, 2013 #19
    If your answer to this is a definite no, preferably with some explanation, it will be one way for me to see the error of my ways :)

    Well this seems to be at the heart of this discussion.
  21. May 9, 2013 #20
    A light cone is the path that a flash of light, emanating from a single event (localized to a single point in space and a single moment in time) and traveling in all directions, would take through spacetime.

    so yes it is possible for a galaxy to be in one persons light cone as opposed to another observer.

    1 example is the light cone from observer "a" sees a galaxy at the edge of the observable universe. However observer "B" on the opposite side of the universe would not see the same galaxy.

    Yes scalar vs vector is the heart of the problem.

    imagine this thought experiment.

    LOL imagine teaching your 5 year old vector time???

    I have a doctors apointment at 3:00coordinate x,y,z by my watch its 2:30coordinate 1x,1y,1z so I only have the vector half an hour to arrive.

    sounds confusing, as well as impractical

    lol imagine teaching your 5 year old vector time?

    edit: here is another thought experiment. I'm driving down a windy road, I have 400 km to go and I'm going 100 km. how long does it take me?
    as a scalar the answer is easy, as a vector I would have to factor in every vector angle along the path of travel.
    Last edited: May 9, 2013
  22. May 10, 2013 #21
    I imagined explaining myself to my wife this way. That was funny enough. But seriously:

    Your example:
    Mine from post 1:
    ...and my example was in relation to the picture:

    It IS the same, and I said also:

    So while I may have problems with proper names and terms, my understanding seems to be quite ok.

    Mordred, do you realize that by saying, I will quote it again:

    you admit that other galaxies actually *must* be under influence which we cannot, and never will be able to see directly? A point I tried to make from post one? Something I have never read about, except when in context of things like unexplainable dark energy?
    Last edited: May 10, 2013
  23. May 10, 2013 #22
    There is the observable portion of the universe. Beyond that obviously galaxies. However the amount of influence from outside the observable on our observable would be no different than what occurs inside the observable universe. So although we cannot directly measure the energy mass density of that region. We can calculate its influence in the same manner we do with the observable universe.
    This fact has nothing to do with time being a scalar property.
    Last edited: May 10, 2013
  24. May 10, 2013 #23
    Lol I keep thinking about how one would go about making a vector time wrist watch. Vectors requiring a direction as well as the scalar value. Each watch would need a compass or GPS lol.
    Then on top of it you would need reference points and location of your destination relative to you. As well as a good calculator lol

    Makes one glad time is scalar and not a vector. Far Far easier.
  25. May 10, 2013 #24


    User Avatar
    Science Advisor

    Note that for a space-time with a space-llike foliation by a one-parameter family ##\Sigma_{t}## and an open subset of it covered by a coordinate chart adapted to the foliation (e.g. Schwarzschild coordinates for the associated space-time) we can interpret the coordinate vector field ##(\frac{\partial }{\partial t})^{a}## as the "time direction" or "flow of time" (you can interpret ##\nabla^{a}t## in the same way since the vector fields will be proportional). This is of course a very specialized situation since general space-times don't admit such coordinate charts / foliations.
  26. May 12, 2013 #25
    Thank you for this attempt with gradients, WannabeNewton, I didn't think about it, but this still implies treating time as scalar. No, the idea as I see it gives time the quality of a vector (or some tensor) and if this is wrong, then go ahead, shoot it and sink it, and let me drown with it because I will still try to defend it (so I deserve going down for stubborness).

    Mordred, I like your humour. I also hope we will never have to deal with vector time wrist watches. And as Wikipedia says:

    meaning that '1 minute', '5 minutes', '10 minutes' is generally enough, and it is enough for all simple purposes of every day life-

    -however, if you posted in Cosmology forum, and you wanted to be really really precise, you would actually need something to diffrentiate between "10 minutes time from where I am now to where I am now" and "10 minutes from here to the Sun". These would be two different things for any third observer. Similarly, there is a difference between 10 minutes in Milky Way and 10 minutes in Bullet Cluster way out there.

    Imagine there is generally only one kind of movement on Earth: free fall; down. Someone suddenly lifted us up and now we are falling. Would you call this movement scalar, with one direction (to the center), or vector, as seen by an observer from the space? In this example, any distance between 2 falling objects will shorten over time. It is better to explain it by understanding the mechanics of gravity, rather than believe in shrinking space. Similarly, I would rather explain redshifts by understanding the mechanics of time, instead of simply believing in expanding of space.

    If you come across a text specifically about time in context of scalar or vector (or something else), I would appreciate a link or a title. I did not find anything very specific, only a number of opinions here and there in the Net, most stating that 'time is scalar', but newer with any serious thought behind them.

    For now, I gave up on the long articles about light cones, but thanks for the links to the shorter pages, they helped with terminology.

    Now that is something I accept as a reasonable argument, even despite the fact that we would be talking about a homogeneous universe with equal mass distribution.
    I didn't know (although I hoped) that such things are considered, and I didn't know (and I still don't) what is the scale of such influence. I imagine it would be greater near the edges of our observable universe, especially in directions where most energy/mass has been found, and smaller near the Milky Way, not only because of the distance, but also because of accleration that happened since there (a mental shortcut :).

    Maybe it has nothing to do with time being scalar, but in my humble opinion, if time were truly scalar, the universe wouldn't need to expand, and we would end up facing the newtonian gravity paradox.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook