Length Contraction

The Geometric Origin

To understand why length contracts, we must first clarify how length is measured. In any inertial frame, the length of an object is the distance between its two ends measured simultaneously in that frame. This is the key distinction from time dilation. For time, we compare two events at the same place in one frame (the proper time). For length, we compare two events (the positions of the front and back of the object) that occur at the same time in the observer's frame.

Consider a rod of proper length \(L_0\) at rest in frame \(S'\). In \(S'\), the coordinates of the back end are \(x'_1 = 0\) and the front end are \(x'_2 = L_0\). These measurements are simultaneous in \(S'\) because the rod is at rest; it does not matter if we measure the ends at slightly different times, they stay at \(0\) and \(L_0\).

Now consider an observer in frame \(S\), where the rod moves with velocity \(v\). To measure the length \(L\) in \(S\), the observer must record the positions of the front and back ends at the same instant \(t\) in frame \(S\). Let these events be \(E_1\) (back end) and \(E_2\) (front end). In frame \(S\), the length is \(L = x_2 - x_1\), where \(t_1 = t_2\). However, because of the relativity of simultaneity, these two events are not simultaneous in the rod's rest frame \(S'\). This mismatch in simultaneity is the geometric source of the contraction.

Deriving the Contraction Factor

We can derive the length contraction formula directly from the Lorentz transformation. Recall the transformation for the spatial coordinate \(x'\) in terms of \(x\) and \(t\): \[ x' = \gamma (x - vt) \] where \(\gamma = \frac{1}{\sqrt{1 - v^2/c^2}}\). Let the back of the rod be at \(x_1\) and the front at \(x_2\) in frame \(S\). We measure these positions simultaneously, so \(t_1 = t_2 = t\). The coordinate of the back end in \(S'\) is: \[ x'_1 = \gamma (x_1 - vt) \] The coordinate of the front end in \(S'\) is: \[ x'_2 = \gamma (x_2 - vt) \] The proper length \(L_0\) is the distance between these ends in the rest frame \(S'\): \[ L_0 = x'_2 - x'_1 \] Substituting the Lorentz expressions: \[ L_0 = \gamma (x_2 - vt) - \gamma (x_1 - vt) \] \[ L_0 = \gamma (x_2 - x_1) \] Since \(L = x_2 - x_1\) is the length measured in \(S\), we have: \[ L_0 = \gamma L \] Solving for \(L\): \[ L = \frac{L_0}{\gamma} \]

Directional Dependence

A critical detail often missed is that length contraction occurs only in the direction of relative motion. If the rod is moving along the \(x\)-axis, its length in the \(x\)-direction contracts, but its dimensions in the \(y\)- and \(z\)-directions remain unchanged. Why? The Lorentz transformation for transverse coordinates is simply \(y' = y\) and \(z' = z\). There is no mixing of space and time in these directions. Therefore, a sphere moving at relativistic speeds does not become a flattened pancake in all directions; it appears as an ellipsoid only if you consider the visual appearance (which involves light travel time delays), but its measured geometric shape is contracted only along the axis of motion. This anisotropy is a direct consequence of the structure of the Lorentz group, which leaves transverse spatial intervals invariant.

Reciprocity and Symmetry

Length contraction is reciprocal. If observer \(A\) sees observer \(B\)'s ruler contracted, observer \(B\) also sees observer \(A\)'s ruler contracted. This seems paradoxical: how can both rulers be shorter than the other? The resolution lies in the relativity of simultaneity. When \(A\) measures \(B\)'s length, they mark the positions of \(B\)'s ruler's ends at the same time in \(A\)'s frame. When \(B\) measures \(A\)'s length, they mark the positions of \(A\)'s ruler's ends at the same time in \(B\)'s frame. These two sets of "simultaneous" events are different sets of events in spacetime. There is no contradiction because the measurements are not being made of the same pair of events. The symmetry of the Lorentz transformation ensures that the physics is consistent for all inertial observers.