Spacetime and Minkowski Diagrams

Before Einstein, space and time were two separate stages: space was the room the play happened in, and time was a clock ticking impartially in the background, the same for everyone. Relativity tears down the wall between them. Because moving observers disagree about lengths and about durations — and disagree in a linked, self-consistent way — the only thing they all agree on is a single four-dimensional arena that mixes the two together. Hermann Minkowski, Einstein's old teacher, put it best in 1908:

"Henceforth space by itself, and time by itself, are doomed to fade away into mere shadows, and only a kind of union of the two will preserve an independent reality." That union is spacetime.

This page is about learning to see spacetime — to draw it, read it, and use a single picture, the Minkowski diagram, to make the strange results of relativity look almost obvious. Once you can read one of these diagrams, time dilation, length contraction and the relativity of simultaneity stop being separate magic tricks and become different views of one geometric object.

Events: the atoms of spacetime

The basic ingredient is an event: something that happens at a definite place and a definite instant. A firecracker going off, two particles colliding, your alarm ringing — each is a point in spacetime, tagged by four numbers (t, x, y, z): when, and where in three directions. An event is not a thing that moves or lasts; it is a single point of "here-and-now".

To draw this on paper we throw away two of the space directions and keep just one, x, against time. But we do not plot t up the page — we plot ct, time multiplied by the speed of light. Why? Because ct is a distance (metres), the same units as x, so the two axes are measured in the same currency. It also makes light travel along tidy 45^\circ lines, as we'll see. By long-standing convention ct runs up the page and x runs across it, so "later" is higher up.

The vertical axis is ct (time, in distance units); the horizontal axis is x (space).
A single point is an event — a "when-and-where".
A line or curve traced out by a moving object is its worldline — its whole history, past to future.

Worldlines: histories drawn as curves

Follow any object through time and it sweeps out a continuous trail of events — its worldline. An object sitting still stays at the same x while ct keeps increasing, so its worldline is a vertical line. An object cruising at constant speed v drifts sideways as time passes, so its worldline is a tilted straight line. Accelerate, and the worldline curves.

Here is the crucial twist. On a Minkowski diagram, the faster something moves, the more its worldline tilts away from vertical. A vertical worldline covers a lot of time and no distance (slow, or at rest); a worldline tilted towards 45^\circ covers a lot of distance in a little time (fast). Since the slope of a worldline is \dfrac{\Delta(ct)}{\Delta x} = \dfrac{c}{v} = \dfrac{1}{\beta} where \beta = v/c, a slower object has a steeper line. Nothing with mass can reach the 45^\circ line — that slope belongs to light alone.

The light cone: the shape of cause and effect

Send a flash of light out from an event. Because light moves at c, in our ct-vs-x picture it travels along lines of slope \pm 1 — perfect 45^\circ diagonals (ct = \pm x). These two lines through an event form its light cone (a genuine cone once you restore the other space directions). The light cone slices spacetime, as seen from that event, into three regions with completely different physical meanings.

Explore the diagram below. It builds up a stationary worldline, a faster tilted worldline, the 45^\circ light cone, the three regions it carves out, and finally the tilted axes a moving observer would draw.

The future (upper wedge, inside the cone): events this event can still reach and influence, since a signal from it need travel no faster than light. These are separated from the event by a timelike interval.
The past (lower wedge, inside the cone): events that could have sent a signal to this one — its possible causes. Also timelike.
Elsewhere (the left and right wedges, outside the cone): events so far away in space and so close in time that not even light can bridge the gap. No cause and effect either way; the separation is spacelike. Different observers can't even agree on their time order.

Why the moving observer's axes tilt

The final step of the diagram is the payoff. A second observer flying past at speed v is perfectly entitled to draw their own spacetime diagram, with their own time axis ct' and space axis x'. Where do those axes go on our picture?

Their ct' axis is the set of events at their own location x' = 0 — that is just their own worldline. It tilts towards the light line, with slope 1/\beta.
Their x' axis is the set of events they call "simultaneous with the origin" (t' = 0). The Lorentz transformation puts this line at slope \beta — tilting up towards the light line by the same angle the time axis tilts across.

Both axes scissor symmetrically towards the 45^\circ line, like a closing pair of shears, keeping the light line as their bisector. This single geometric fact explains everything: because the moving observer's x' axis is tilted, the events they call simultaneous are not the ones we call simultaneous — that is the relativity of simultaneity, staring at you from the page. Time dilation and length contraction are just the mismatch between the tick-marks along the tilted axes and the straight ones.

Here's the beauty of the scissoring. The ct' axis tilts clockwise towards the 45^\circ line, and the x' axis tilts anticlockwise towards it by exactly the same angle. So the light line stays exactly halfway between the moving observer's two axes — which means, in their coordinates too, light still has slope 1, still travels at c. The whole apparatus of tilted axes is engineered precisely so that every observer measures the same speed of light. The diagram isn't just illustrating relativity's second postulate — it's built out of it.

The moving observer's axes are not perpendicular on our diagram — and that is fine. Newcomers see the tilted ct' and x' axes making an acute angle and think something has gone wrong, or try to "measure" distances on the diagram with an ordinary ruler. Two traps to avoid:

Don't judge lengths and angles by Euclidean eyeballing. Spacetime geometry uses the Minkowski "distance" s^2 = (c\Delta t)^2 - \Delta x^2, with a minus sign, not Pythagoras. On the page a worldline that "looks longer" can actually be a shorter proper time. The tick-marks along the tilted axes are stretched compared with ours; you can't read them off with a straight ruler.
The tilt is not a perspective drawing. Neither observer is "really" using bent axes — each sees their own axes as a perfectly ordinary square grid. The tilt only appears when we insist on drawing their coordinates on top of ours.

Worked example: is this cause or coincidence?

Two firecrackers go off. In our frame, firecracker A is at (ct_A, x_A) = (0, 0) and firecracker B is at (ct_B, x_B) = (2, 5) (in light-seconds). Could A have caused B?

A could influence B only if a signal from A could reach B — that is, only if B lies inside or on A's future light cone. Compare the time gap and the space gap:

c\,\Delta t = 2 \text{ light-seconds}, \qquad \Delta x = 5 \text{ light-seconds}.

The spatial separation 5 exceeds the light-travel budget 2, so even light couldn't get from A to B in time. Equivalently, the interval is

s^2 = (c\Delta t)^2 - \Delta x^2 = 2^2 - 5^2 = 4 - 25 = -21 < 0,

which is spacelike: B is in A's "elsewhere". A cannot have caused B; the two events are causally disconnected, and some observers will even see B happen before A. Had the numbers been (6, 5) instead, we'd get s^2 = 36 - 25 = +11 > 0, a timelike interval with B safely inside A's future cone — cause and effect back in business, and every observer agreeing A came first.