Why is the speed limit of the universe tied to light and not to gravity or something else? Does light really travel at that limit? And what if the photon turned out to have a tiny mass? These questions intrigue me, and they turn out to share one answer.
It starts with an idea: c isn’t really about light. It’s the speed limit on cause and effect, the fastest anything can travel from one event to another and still change what happens there. Light just happens to be able to reach it. I’ll build up to that from something software engineers already know.
Why two events need a path between them
Say a server writes a value. When can that write affect something happening somewhere else? Only if something can travel from the first place to the second: a packet, a push, a signal. If nothing can get there in time, the write can’t change what happens over there, and whatever happens over there can’t change the write.
Distributed-systems people have a name for something close to this, an idea Leslie Lamport laid out in 1978 using the same kind of space-time diagram that relativity does. Two events are concurrent when neither happened-before the other, meaning no chain of messages connects them. Physics asks the same question about the whole universe: is there a maximum speed for those messages? There is, and we call it c.
Drawing the limit
Put event A in the middle of a diagram where space runs sideways and time runs up. The fastest anything leaving A can travel traces two lines, one going left and one going right. Everything between those lines, above A, is what A can affect. With more than one direction of space that region is a cone, which is why physicists call it the future light cone. The mirror image below A is everything that could have affected A: its past. Everything outside the cone is “elsewhere,” events that can’t affect A and that A can’t affect.
Click anywhere on the diagram to move B. If B lands in the upper cone, A can be the cause and B the effect. If it lands in the lower cone, it’s in A’s past, so the influence can only go the other way: B can be the cause and A the effect. Anywhere else, neither can influence the other. The orange lines are the fastest path anything can take.
Light happens to ride those lines, but the diagram itself is only about cause and effect, and the slope of the cone’s edge is c. Every observer moving at a steady speed draws the same cone. That’s the heart of relativity, and it’s what a later post is about.
Why light sits on the edge
Mass is the reason. Anything with mass can get close to the edge but never reach it. Anything without mass has no choice: in empty space it travels exactly on the edge, and it can never slow down or sit still. (A massless thing sitting still would have no energy, so there would be nothing left.) Light has no mass (as far as we can tell; more on that below), so it rides the edge. That’s how c ended up with the name “the speed of light.” People measured light’s speed long before anyone understood that the number was about something bigger.
Which also answers my question about gravity. Gravity isn’t a rival speed limit; gravitational waves are massless too, so they ride the same edge. We got a direct check in 2017, when two neutron stars merged about 130 million light-years away. The gravitational waves arrived first, and a burst of gamma rays followed about 1.7 seconds later. That isn’t because light is slower. The waves come from the merger itself, while the gamma rays most likely come from a jet that takes a moment to form and break out of the debris. Even if the entire 1.7 seconds were a speed difference, over 130 million years it would be at most a few parts in 10¹⁶. The published analysis (LIGO, Virgo, Fermi and INTEGRAL, in The Astrophysical Journal Letters) is more cautious: it assumes the nearest distance the data allows and lets the gamma rays leave up to 10 seconds late, and still keeps any difference within a few parts in 10¹⁵.
Why there has to be a limit
Suppose you could send a signal from A to something “elsewhere.” Observers moving at different constant speeds disagree about the order of events that are elsewhere from each other (a later post is about exactly that). So some observers would see the signal arrive before it was sent. Add a second signal going back and you can get a reply before the question. That’s a race condition no locking can fix, because the disagreement is built into the geometry.
So I don’t see the limit as an engineering constraint we might work around someday. As long as the laws of physics look the same to every observer moving at constant speed, which is where relativity starts, it’s what keeps cause before effect for everyone.
Does light really travel at the limit?
As far as anyone can measure, yes. The Standard Model says the photon is exactly massless, protected by the gauge symmetry behind electromagnetism. Experiments can’t prove zero, but they can bound it. Particle physicists quote mass in energy units (an electron is about 511,000 eV), and the Particle Data Group’s photon listing (2021) gives an upper limit of 1 × 10⁻¹⁸ eV, from how the Sun’s magnetic field behaves out to the orbit of Pluto. That makes a photon more than 10²³ times lighter than an electron, if it has any mass at all.
If the photon did have a tiny mass, light would travel a hair below c, and different colors would travel at slightly different speeds. We don’t see that. And even if we did, c would still be the speed limit of causality, with light falling a hair short of it. That’s the real answer to my “what if”: the limit doesn’t depend on the photon being massless.
One question I can’t leave alone
In distributed systems we’re always trying to sync clocks across machines. In theory, is that even possible? The light cone hints that the problem goes deeper than engineering: how much time a clock counts depends on how it moves, and two observers can disagree about which events happened at the same time. That’s what the next posts are about.