Can Two Observers Disagree About Which Event Happened First?

Suppose two events happen far apart.

At event A, a detector clicks. At event B, another detector clicks somewhere else.

One observer says A happened first.

Another observer, moving relative to the first, says B happened first.

Neither observer made a mistake.

That sounds dangerous. If relativity allows people to disagree about which event came first, what stops one observer from seeing an effect occur before its cause?

The answer is that relativity does not treat every pair of events the same way.

For events that could causally influence each other, their order is protected. For events that are spacelike-separated, their order does not have to be.

That is not a loophole in causality. It is part of the structure that preserves it.

1. Begin Outside the Light Cone

Imagine events A and B on a spacetime diagram.

They are separated by enough distance, and little enough time, that even light could not travel from one to the other before both have occurred.

In other words, they are spacelike-separated.

Neither event can have caused the other. A light signal leaving A would reach B's location too late. A light signal leaving B would likewise reach A's location too late.

The light cone already tells us that much.

But now ask a different question:

Which event happened first?

It is tempting to assume that the universe must contain one correct answer, even if the two events cannot affect each other.

Special relativity says otherwise.

2. Different Observers Divide Spacetime Differently

On an ordinary graph, a horizontal line can mean “all the points at the same height.”

On a spacetime diagram, a line through separated events can play a similar role: it can represent events that an observer regards as happening at the same time.

The surprising part is that different inertial observers do not agree on which distant events belong to the same moment.

For one observer, imagine a slice of simultaneity that places A below B on the time direction. That observer assigns A the earlier time.

For another observer moving relative to the first, the corresponding simultaneity slices are tilted differently through spacetime. In that coordinate system, B can receive the earlier time instead.

And between those two cases there is a frame in which A and B are simultaneous.

Nothing happened to the events themselves. Nobody watched an event run backward. The observers are using different, equally valid ways of assigning space and time coordinates to the same two points in spacetime.

For a spacelike-separated pair, those coordinate systems need not agree about the sign of “B's time minus A's time.”

That is the relativity of temporal order.

3. The Equation Shows Where the Flip Comes From

The geometry carries most of the idea, but one piece of the Lorentz transformation makes the mechanism unusually clear.

If two events have spatial separation \(\Delta x\) and time separation \(\Delta t\) in one inertial frame, another frame moving at speed \(v\) along that spatial direction assigns them a time separation

$$ \Delta t'=\gamma\left(\Delta t-\frac{v\Delta x}{c^2}\right), $$

where \(\gamma\) is always positive.

So the important part for event ordering is what sits inside the parentheses:

$$ \Delta t-\frac{v\Delta x}{c^2}. $$

The first observer may have \(\Delta t>0\), meaning B happens after A.

But if the events are far enough apart in space compared with their separation in time, changing \(v\) can make that expression zero. In that frame, the events are simultaneous.

Change the frame a little further, and the expression can become negative.

Now B happens before A according to that observer's coordinates.

This possibility exists precisely for spacelike separation. The large spatial separation gives the second term enough leverage to change the sign of the time difference while the observer still moves slower than light.

The equation is not creating the effect. It is simply exposing in algebra what the spacetime diagram already contains.

4. Why Nothing Has Gone Wrong

Suppose observer 1 says

A first, then B.

Observer 2 says

B first, then A.

At first glance, that sounds like a disagreement about causality.

But remember what kind of events A and B are.

They are spacelike-separated.

A cannot send a light-speed-or-slower signal to B in time to influence it. B cannot send one to A either. There is therefore no physically allowed chain of cause and effect running from one of these particular events to the other.

Their ordering can change because there is no causal ordering between them to preserve.

Relativity is not saying that cause and effect depend on your point of view. It is saying that the words before and after, when applied to sufficiently distant events that cannot affect one another, contain more observer-dependence than everyday intuition suggests.

That distinction matters.

5. Causally Connected Events Are Different

Now move B inside A's future light cone.

Enough time passes between the events that a physical influence traveling at or below the speed of light could get from A to B.

The events are now timelike-separated. If B lies exactly on the boundary, they are lightlike-separated.

Observers can still disagree about plenty of details. They can assign different coordinate times to A and B. They can disagree about how much time separates them.

But they cannot reverse their order.

If B lies in A's causal future, every inertial observer agrees that A occurs before B.

There is no allowed change of inertial frame that turns

A → B

into

B → A.

That is the protected structure.

The same relativity that allows the ordering of spacelike events to change refuses to reverse the ordering of events that could stand in a causal relationship.

6. What Relativity Actually Gives Up

The strange conclusion is therefore narrower than it first appears.

Relativity does not say that all chronology is subjective.

It says that the universe does not provide an observer-independent chronological ordering for every pair of distant events.

For spacelike-separated events:

A before B, B before A, and A simultaneous with B can all be valid descriptions in different inertial frames.

For timelike- or lightlike-separated events, the direction of causal order remains fixed.

So the question

Which happened first?

has an observer-independent answer when the relationship between the events permits a causal ordering.

When it does not, nature does not insist on supplying one.

That is why two observers can disagree about temporal order without disagreeing about cause and effect.

Relativity gives up universal chronology.

It does not give up causal consistency.

---

If the spacelike/timelike distinction itself still feels slippery, the natural next step is The Spacetime Interval: Timelike, Lightlike, and Spacelike Separation. For the geometric picture behind the entire distinction, return to Light Cones: What Can Affect What in Spacetime?

And there is a sharper question waiting just beyond this one: if spacelike event order can reverse between frames, what would happen if information really could travel between spacelike-separated events? That is where Why Faster-Than-Light Communication Threatens Causality begins.