Two distant events can be simultaneous in one inertial frame and occur at different coordinate times in another. This does not mean that an observer changes the events by looking at them. It means that assigning a time to a distant event requires a specified comparison procedure.
Begin with an operational definition
An observer can send a light pulse to a distant reflector and receive its return. If the local emission and reception readings are t₋ and t₊, Einstein radar time assigns the reflection event the midpoint (t₊ + t₋)/2. The corresponding radar distance is c(t₊ − t₋)/2.
This construction uses a clock, a signal and a synchronization convention. It distinguishes the time assigned to a distant event from the later time at which information about that event is received. The book develops this distinction in clocks and radar coordinates.
Different inertial frames using Einstein synchronization do not share the same surfaces of equal time. Their coordinate assignments are connected by Lorentz transformations. Agreement about local meetings and causal relationships survives this difference.
Transform the same pair of events
Choose a length L and express time in units of L/c. Event A is at x = 0, t = 0. Event B is at x/L = 1 and ct/L = δ. A frame moving at speed βc along the positive x direction assigns
cΔt′/L = γ(δ − β), Δx′/L = γ(1 − βδ), γ = 1/√(1 − β²).
The interval in units of L² is δ² − 1 in both frames. The moving frame calls the events simultaneous when β = δ, which is physically available only if |δ| < 1. Thus only spacelike separated events can be made simultaneous by a subluminal inertial boost.
Change the comparison frame
TWO EVENTS · FLAT SPACETIME · EINSTEIN SYNCHRONIZATION
Diagram uses the original frame's x/L and ct/L coordinates. Green: t′ = 0. Grey dashed lines: light rays through A. Event B stays at x/L = 1.
The separation is spacelike. This moving frame assigns B an earlier time than A; no light signal can connect the two events.
Which order can change?
With δ = 0 and β = 0.6, the moving frame obtains cΔt′/L = −0.75. Reversing the frame velocity changes this to +0.75. The same two events have different coordinate orderings, but remain spacelike separated.
Now set δ = 1.5. For every allowed boost, δ − β stays positive. Event B lies in A's timelike future, and every future-oriented inertial frame agrees that B occurs later. At δ = 1 the separation is null and the future order is also preserved.
The relevant distinction is causal structure, not whether a signal happens to arrive first at a chosen telescope. Reception order includes travel distances. The demonstration transforms emission-event coordinates; it does not simulate their visual appearance.
Where an inner ordering could enter
SMM/DTT may distinguish inner generation from completed observations. A proposed inner sequence need not equal the simultaneity convention used by every observer. The book's discussion of order and physical clocks explicitly separates these roles.
If the source ordering has observable consequences, the model must calculate them and confront relativistic tests. If it is descriptive freedom, exchanging independent source steps must not alter accessible records. Either possibility requires more than identifying one coordinate t with an absolute cosmic present.
This gives readers a focused question: does a proposed chronology change a measured correlation, or does it reorganize the description of an unchanged experiment? The answer determines whether there is a new physical prediction to test.
Sources and further reading
- Feynman, Leighton and Sands, Space-Time, for Lorentz transformations and simultaneity. The event-pair demonstration is an original teaching calculation.
- Continue with light cones and relativity and becoming.

