The Reality of Time

The idea that time isn’t “real” is an ancient one — if we’re allowed to refer to things as “ancient” under the supposition that time isn’t real. You will recall the humorous debate we had at our Setting Time Aright conference a few years ago, in which Julian Barbour (the world’s most famous living exponent of the view that time isn’t real) and Tim Maudlin (who believes strongly that time is real, and central) were game enough to argue each other’s position, rather than their own. Confusingly, they were both quite convincing.

smithsonian-mag The subject has come up once again with two new books by Lee Smolin: Time Reborn, all by himself, and The Singular Universe and the Reality of Time, with philosopher Roberto Mangabeira Unger. This new attention prompted me to write a short essay for Smithsonian magazine, laying out the different possibilities.

Personally I think that the whole issue is being framed in a slightly misleading way. (Indeed, this mistaken framing caused me to believe at first that Lee and I were in agreement, until his book actually came out.) The stance of Maudlin and Smolin and others isn’t merely that time is “real,” in the sense that it exists and plays a useful role in how we talk about the world. They want to say something more: that the passage of time is real. That is, that time is more than simply a label on different moments in the history of the universe, all of which are independently pretty much equal. They want to attribute “reality” to the idea of the universe coming into being, moment by moment.


Such a picture — corresponding roughly to the “possibilism” option in the picture above, although I won’t vouch that any of these people would describe their own views that way — is to be contrasted with the “eternalist” picture of the universe that has been growing in popularity ever since Laplace introduced his Demon. This is the view, in the eyes of many, that is straightforwardly suggested by our best understanding of the laws of physics, which don’t seem to play favorites among different moments of time.

According to eternalism, the apparent “flow” of time from past to future is indeed an illusion, even if the time coordinate in our equations is perfectly real. There is an apparent asymmetry between the past and future (many such asymmetries, really), but that can be traced to the simple fact that the entropy of the universe was very low near the Big Bang — the Past Hypothesis. That’s an empirical feature of the configuration of stuff in the universe, not a defining property of the nature of time itself.

Personally, I find the eternalist block-universe view to be perfectly acceptable, so I think that these folks are working hard to tackle a problem that has already been solved. There are more than enough problems that haven’t been solved to occupy my life for the rest of its natural span of time (as it were), so I’m going to concentrate on those. But who knows? If someone could follow this trail and be led to a truly revolutionary and successful picture of how the universe works, that would be pretty awesome.

  1. Paul, thanks for your comments. A few clarifications:-
    1. You’re partial to the first definition of time…well that’s a good start. But, sorry, you can’t pick and chose. To only “take” one definition means you only have half an idea of the whole that the word Time implies. The problem people have with understanding Time is that they don’t differentiate the two distinct meanings…and invariably they mix and confuse the two. Hence they go around in ever decreasing circles. For a theorist not to state at the outset of a discussion which of the two meanings he is referring to (Smolin etc never do) immediately indicates he doesn’t know his starting point, and hence his speculation is bound to be flawed.

    2. When Sean says time exists, that implies a tangible concrete reality. I am saying it is demonstrably NOT a concrete tangible reality. Its an abstract in both its core, root definitions. Its an abstract framework (which you like), or its a collective noun, a non-specific set of events (which you don’t, yet, care for). And it is hence derived, and is derived from the underlying fundamental, which is ‘events’.
    Let me expand on these two definitions slightly. Definition one: another example of an abstract framework might be the metric system – its not a tangible reality.
    And definition two: another example of a non-specific set (a mass noun) would be ‘fashion’. Like time, fashion is not a tangible reality, its abstract (although the clothes, for example, it refers to might be tangible – for they are the underlying fundamental).
    Does that help??

  2. The block universe conception of time has immediate implications for free will, since all future moments (like past and present moments) are fixed in four dimensional spacetime. Our choices and actions participate in the static law-like patterns (world tubes) that we experience from our time-bound perspective as sequential cause and effect, but it isn’t as if we can stand apart from the pattern and make it other than what it is. So the block universe picture is more or less a timeless, time-neutral depiction of determinism, plus whatever randomness might exist.

    Just as Einstein thought time as we ordinarily think of it as an illusion, he was of the same mind about (libertarian, contra-causal) free will. As is Brian Greene in the Fabric of the Cosmos. Which is fine, since were we somehow not embedded in spacetime, we’d have no basis for our choices and no causal powers,

  3. Magnema, thanks.

    simon morley,
    I think we’re not really disagreeing about point 1. Sure, right now, people use the word time with (at least!) the two meanings you say. I make a recommendation, hey folks, let’s clear up future potential confusion by preferring to think about time this one way.

    As for tangible concrete realities – well, I think there is an integer between 2 and 4. I also think there is a time called 5:04pm EDT, April 6 2015. I’m not sure just how concrete and tangible this latter commitment of mine is. Although, there are a lot more empirical tests I could do that would tell me something about the latter statement.

    Tom Clark,
    People have cognitive illusions about time (that it has an objective “flow”) and about free will (that it is contra-causal). They also have cognitive illusions about whales (that they are fish) and life (that it is powered by elan vital). None of this tells us much about whether the concepts in question succeed or fail to refer. Except … B-theory of time, combined with CPT-invariance of laws, pretty much annihilates a key premise (“fixity of the past”) of the most prominent argument for incompatibilism about free will and determinism.

  4. @paul torek. Sadly, Paul, integers are abstract (not tangible). And the time 5.04 is also abstract (its merely a reference point in our abstract framework). So you have no empirical evidence of time being tangible. Which is what I said. (you do know the difference between abstract and tangible??)
    And you can’t pick and choose which definition you use, even by invoking the support of “hey folks” – you clearly can’t see the significance of the difference. They are two separate meanings. Which one are you hypothesising about??

    Oh, and by the way, what I have shown you is how to define time without reference to itself. Which is why it becomes, fundamentally, redundant….

  5. Speaking as a layperson whose cognitive peace of mind is destroyed by this disorderly topic, I will hope for an eventual Copernicus-Kepler “moment” in the science community to provide a provable platform from which to proceed.

  6. @Jean VL
    Sadly your “moment” won’t come from the science community. “What is Time?” is not a science question…its about semantics. Define the word first and foremost…basic academic rigour that has been ignored on this topic since way before Einstein….including by our own Prof Carroll.

  7. simon morley,

    I explained poorly regarding the two definitions of “time”. My bad. I do see the significance of the difference, which is precisely why I want to stop using the same word for both. Something’s gotta give. Since the t in physics is better-known and more versatile, I propose that the new terminology, say “set of events”, should be coined for the set-of-events concept. As an added bonus, future physics students will still understand why “time” is designated by t.

    3 is an integer, and exists (thus “there exists an integer between 2 and 4” is true), and as you so correctly point out, 3 is not tangible. Therefore “exists” does not imply “is tangible”. Therefore if you want to accuse Sean Carroll of implying time is tangible, you’ll need more than that he says time exists. (Actually, that’s another my bad – he didn’t say “exists”, he said time is real. In my book, “is real” implies “exists”, but this is philosophical territory, so I probably shouldn’t assume he agrees.)

  8. Paul
    Hope you don’t mind me saying, but I feel you’re deliberately skating here, this is important.
    “Is real” as opposed to “isn’t real” infers tangible noun v abstract noun. We know the word exists, therefore the whole debate is about what the word stems from…tangible or not tangible. If it stems from not tangible, i.e abstract, then there is no further need to debate the nature of time. You have solved what Scientific America calls the Ultimate Paradox. For if time is abstract (and we all are agreeing, in the absence of any evidence to the contrary, that it is) then its definition is merely what we ascribe to it (by usage).
    So, to conclude from this clear position:-
    Firstly, the t in science needs no further debate, we agree. It is an abstract framework, and hence comprehensively understood….just as we understand the metric system. Done and dusted. Nothing more to debate.
    Secondly, despite what you say about ignoring the “set of events” definition, its actually this concept that (for some bizarre reason) scientists struggle with. Crack it and you will put out of your mind once and for all any notion that the nature of time requires any more debate; its now wholly understood…right here in this paragraph:- Time doesn’t ‘pass’, only events happen. For you to be able to assert that time ‘passes’, you would need to be able to say conclusively what time is – else you cant make assertions about it. As you can’t do that (i.e. say what time is other than the abstract framework definition above), all you can therefore assert is that events happen – i.e. events are ultimately all we ever witness (e.g. the event series of the earth spinning, or a clock ticking).
    So time is 1. an abstract framework or 2. a set of events. Hey presto, The Ultimate Paradox solved!
    Hence time ceases to be anything but a shorthand word…useful but fundamentally unimportant.
    So why do the scientific community still try to debate the nature of time…you and I have cracked it Paul!

  9. I think the real question Sean poses is not whether or not time is real, but what resources should we expend exploring that particular potential illusion. I’m with Sean on that one. Not many. Aside from the other questions in physics, biology and social science questions are taunting us. Time can wait.

  10. I love this subject, and I’ve written quite a bit about it which you can see on my website. It especially answers some questions raised by Ben Goren, about why we perceive time to run in one direction, why there is an arrow of time etc.

  11. Henry K. O. Norman, thank you for your kind reference to my essays, and thank you for your convenient anthology. I hope the ideas will speak for themselves to anyone willing to read them with an open mind.

    Despite Mr. Carroll’s assertions to the contrary, I believe that the “block universe” already is well on its way to the dustbin of science history. We can retain the empirically validated parts of relativity while jettisoning its needless, unhelpful, metaphysical “block universe” baggage, as suggested in the recent Unger/Smolin book as well as in my own essays to which you referred above in your own post.

  12. . C. N. Smith commented on The Reality of Time.

    Despite Mr. Carroll’s assertions to the contrary, I believe that the “block universe” already is well on its way to the dustbin of science history. We can retain the empirically validated parts of relativity while jettisoning its needless, unhelpful, metaphysical “block universe” baggage, as suggested in the recent Unger/Smolin book as well as in my own essays to which you referred above in your own post.


    That’s Dr Carroll BTW. You may have jettisoned the block universe in your own mind, but there are lots of other minds who would disagree. You can have your own private history I guess but you shouldn’t presume to speak for the whole physics community. Dr Smolin’s views are , in my opinion , outside the general consensus of the physics community, that doesn’t necessarily make him wrong, but it does mean his views are a poor choice to base a definitive science “history” on.

  13. According to Special Relativity, an astronaut orbits a Black Hole for 1 year, then returns home to find that 100 years have passed on the Earth. The “passage” of time differed for the astronaut compared to Earthers. How does “presentism” fit here? I suppose we can say that time was merely stretched out for the astronaut, but I think this also requires that time be continuous and rules out discrete “packets” of time. The “present” would have to be sliced up much finer for the astronaut, and this then contradicts the “presentism” model.

  14. Guess I shouldn’t post comments before morning coffee. The point I was trying to make is that Presentism doesn’t seem to be compatible with Special Relativity’s Time Dilation, particularly if time is not continuous but exists as discrete units at the Plank scale. Let’s say at present I meet an astronaut who journeyed around a black hole. I previous met this astronaut before his journey, 20 years ago by my reckoning but only 10 years ago by his reckoning. If we shared the Present both then and now, and at all times in-between, then each slice or interval of “now time” occurs over a shorter period or interval of time for me than for him. But this breaks the Law of Equivalence. In fact if time had slowed enough for him, then our common “now” slices could occur for me on a time scale smaller than the Plank time scale. I think this is a problem for the Presentism interpretation of time.

  15. Jon Harris,
    Your website also answered some questions of mine, thanks. The link at your name does not work, though: the final e in universe is missing.

    Question: if you like the Everett interpretation, why do you say we no longer have a fully deterministic future and past? The whole quantum multiworld(? not sure what language to use) evolves determinstically, no? It’s just that a “single” me of the future might remember the earlier state and say, “huh, the electron went through slit 1, I couldn’t have predicted that.” Yet at the same time a “different” me, decohered from the first, is saying, huh, the electron went through slit 2.

  16. Bob Zannelli commented, “. . . you shouldn’t presume to speak for the whole physics community.”

    Of course not, and I did not presume to do so when I commented that, “I believe that the ‘block universe’ already is well on its way to the dustbin of science history.” That is my own opinion, albeit one apparently shared by at least a few others.

    I believe that the fundamental problem with the block universe can be traced back to the operational definition of time (time is that which is measured by clocks) upon which the notion of the block universe ultimately rests, via relativity. While the operational definition of time is a perfect fine and useful definition — which, by the way, does an admirable job of supporting the empirically validated parts of relativity — I contend that the operational definition simply is not sufficiently robust to support all of the added metaphysical baggage that has been thrust upon it in the form of the block universe interpretation.

    A relatively brief essay outlining the reasons for my contention, ‘Rethinking a Key Assumption About the Nature of Time,’ may be found here: .

  17. JCN
    You forsee problems due to “operational definition of time (time is that which is measured by clocks)”
    I have been making this point in other posts here. The word Time has TWO distinct uses (definitions) and “that which is measured by clocks” is demonstrably incomplete.

    What I do not understand is why in this subject of Time – above ALL other subjects – the basics of definition have been ignored since time began (pun intended!)

    If one can’t define what it is one is trying to hypothesize about fully, the hypotheses are hence invariably, at best, dodgy.

  18. It would seem like possibilism would be the correct view of time, if quantum mechanics was truly based on probabilities. Then it would mean that only time travel to the past was possible and not the future, if someone was in the “true present” and could avoid the temptation of killing their grandfather. By the true present I mean the present at the top of the graph. If someone traveled back in time, they might feel like they were in the present, but they would actually be at another location in the graph and they would be almost just as accurate as describing their present as they would in the graph of enternalism. Then eternalism would allow full access to the past and future at any point along the graph (again if you could avoid killing your grandfather), but presentism wouldn’t allow for time travel at all. I find that unlikely, since particles can influence each other backwards in time.

    It seems like a time machine could only really rule out presentism, if it the models were going to be tested. Then only an inability to travel to the future at all would rule out eternalism. Then if travel to the past and future was possible, it could never really rule out possibilism with absolute certainty. That would most likely be the case if it wasn’t the first time machine used throughout all of time, and there was never a single time traveler throughout all of history, known or otherwise.

    It would seem like self fulfilling paradox’s would be rather strange in a possibilistic universe. Originally, the time traveler would only be a statistical representation of how he/she would most likely be. He/she wouldn’t actually exist, because his/her time frame wouldn’t have occurred yet. Then someone could be visited by different changing “manifestations” by there being a high probability that they would be visited by a time traveler in the future, and they changed the probability of how those events would occur.

  19. @John Barrett
    With respect John, the idea that you can “travel back in time”, given that the past (universal quantum configuration) is erased (there is only ever one universal configuration at a time), would require you to re-assemble an entire prior universal quantum configuration…that would require you to not only know said configuration (of every quantum particle), but be able to re-arrange every quantum particle back into its prior place.
    Hard as this is, it would also mean that as you, the time traveller, have now appeared in this re-assembled configuration, you have contaminated it…it is no longer the same universal configuration…it is no longer the same prior “moment in time”.

  20. Simon,
    I agree with you about the importance of defining terms. Elsewhere, I wrote,

    ” . . . choose at random virtually any book or essay on the nature of time and search in it for the author’s definition of what he or she means by the word “time.” Be forewarned, however; you would be well advised not to hold your breath until you find it. Countless books have been written about the nature of time, time travel, the philosophy of time, and the whole gamut of related topics, without any effort having been made by the author to define exactly what is meant by the word time.

    To be charitable, perhaps these authors simply assume, or hope, that everyone will already know what is meant by the word time without any need of further clarification, but this is exactly the problem; everyone does not know what time is. If they did, a great many of these works would not be needed. One of the most frequently repeated quotations in the literature of time is one attributed to Saint Augustine of Hippo: “What then is time? If no one asks me, I know: if I wish to explain it to one that asketh, I know not.” With all due respect to Saint Augustine, this simply will no longer suffice; we must do better.”

    For more about this, please read my essays.

  21. JCN
    It’s a great pleasure to read your essay. Finally I have found someone who is on the same page as me in this minefield (to mix my metaphors)!
    I haven’t read John Archibald Wheeler, though I intend now to do so…but his quote that you use “We don’t realize we’re the source of the puzzle, because we invented the word ” is precisely the view I took when putting forward my own definition.
    I am at a loss to understand why so many academics…both in science, and more unforgivably in philosophy… just don’t get this. I put this idea to this very blog a couple of years ago and the response was essentially ‘that there are so many languages that it is inconceivable that they would all mis-define time. One language would have surely defined it correctly’.
    Except we are all using a language that doesn’t….!

    In my own essay, I have very deliberately (systematically) reduced the uses of the word time to its core meaning…and this is where, I believe, the paradox lies. Time has two root meanings, and they are used interchangeably because people don’t recognise the subtle but distinct difference in the two. And hence all manner of misadventures occur! [This double meaning might allow for the release of the paradox that time measures time (sort of) in that time (definition 1) calibrates time (definition 2).]

    The other line I take is that if there is no evidence to support the idea that time is a tangible ‘thing’ (and I don’t believe there is) then time is abstract (clearly this is not a unique view). But this does lead to the conclusion therefore, if it is abstract, that NO assertions can be made about it UNTIL it is comprehensively defined ( assertions such as “time passes” or “arrow of time” or “time bends”). You can’t make any assertion about time (or anything else) if you don’t know (precisely) what it is you are asserting about.
    ALL you can assert is what you see…which is invariably ONLY events and change (you never witness time, only events). They then must form the basis of the definition of time. I think you have similar ideas.

    Anyway, I will spend more time in re-reading your essay.

    Mine is here if you’re interested (and yes, I stand by my bold stance in this title in that I believe I have defined time completely!)

  22. Simon,
    Thank you for your kind words. It’s gratifying to find someone else who cares about clearly defining terminology before waxing eloquent on a given topic.

    If you like ‘Toward a Helpful Paradigm for the Nature of Time,’ please don’t overlook ‘Time: Illusion and Reality,’ which may be found here: . Like the paradigm essay, it emphasizes the crucial importance of clarity when defining exactly what it is that we’re talking and writing about with regard to the nature of time. Disregard of this imperative goes far toward explaining why physics currently finds itself in such a needless muddle, in my opinion.

    I will be taking a look at your website and essay. We are all on a quest for the truth and for clarity about the nature of this amazing universe in which we find ourselves; these are precious commodities (truth and clarity) upon which none of us holds a monopoly. Fortunately, we can all learn from each other, a fact which explains how we already have come so far.

  23. @ Simon Morley

    You are assuming that every event in the past is tied to every single event in the present. Say for instance the Mars rover died, then it laid dormant for years afterwards. The rover would no longer have any impact on the events on Earth, since it could no longer transmit anything. Then there would be nothing stopping someone from traveling back in time to recover the rover. If the rover was there or not there on Mars, it wouldn’t have had any impact on the history of Earth. Nobody would have ever missed it!

    Your also assuming that a timeline could not branch or split. It could be possible that, if an event in your history is changed, the timeline or universe would split into another alternate timeline or universe. Then I agree with you that your own eigenstate, for a better choice of words, cannot be altered for a macroscopic object. It would seem like the Grandfather Paradox should include everything that influenced your own eigenstate. That is why I believe that alternate worlds would have to be created in this fashion, and they are not actually created in the way the MWI describes.

    In a self fulfilling temporal paradox, the eigenstate of the time traveler is not changed. The time traveler finds that the events that took place in their own history are the same events that took place from his/her own influences in the past. Then the only way that sort of event could take place, in a possibilistic universe, is if an event could be influenced by there just being a high probability that someone traveled back in time from the future that has not occurred yet, from any time-frame.