In math and physics, the idea of orientation entanglement is sometimes used to help explain the shape and structure of spinors. It also shows why certain types of mathematical groups, called special orthogonal groups, are not simply connected.
Elementary description
Spatial vectors by themselves are not enough to describe how objects rotate in space.
Imagine a coffee cup hanging in a room by two elastic rubber bands attached to the walls. If the cup is turned so its handle makes a full 360° twist around the center of the cup, the handle returns to its starting position. However, the cup’s orientation is now twisted compared to the room’s walls. If the cup is lowered to the floor, the rubber bands form a double helix with one full twist. This shows orientation entanglement: the cup’s new position is different from its original position, even though it looks the same. The twisting rubber bands prove the cup is still connected to the room in a twisted way.
Spatial vectors alone cannot explain this twist. If a vector is drawn on the cup, a full rotation would move the vector back to its original direction. The vector does not show the twist between the cup and the room.
The cup is tightly twisted and cannot be untwisted without moving it again. However, if the cup is rotated twice—720° total—the rubber bands form two full twists. If the cup is then moved through the center of one twist and shifted to the other side, the twist disappears. The bands are no longer twisted, even though no extra rotation was done. (This is easier to test with a ribbon or belt.)
After a 360° rotation, the cup is twisted compared to the room. After a 720° rotation, it is not. A vector alone cannot tell the difference between these two situations. Only when a spinor is attached to the cup can the difference be seen.
A spinor is like a vector with a special feature. It can be shown as a vector with a flag on one side of a Möbius strip, pointing inward. If the cup is rotated 360°, the spinor returns to its starting point, but the flag is now on the opposite side of the strip, pointing outward. Another 360° rotation is needed to return the flag to its original position.
More information about this concept and its connection to math can be found in the article on tangloids.
Formal details
In three dimensions, the problem described earlier relates to the fact that the Lie group SO(3), which represents all possible rotations in three-dimensional space, is not simply connected. To address this, mathematicians use another group called SU(2), which is also known as the spin group in three-dimensional space. SU(2) is a double cover of SO(3), meaning that each rotation in SO(3) corresponds to two distinct elements in SU(2).
If X = (x₁, x₂, x₃) is a vector in three-dimensional space, it can be represented as a 2×2 matrix with complex numbers as entries. The determinant of this matrix, when multiplied by -1, equals the square of the vector’s length in Euclidean space. Additionally, this matrix is trace-free, meaning the sum of its diagonal elements is zero, and it is Hermitian, a type of matrix that remains unchanged when its complex conjugate is taken.
The unitary group, which includes all matrices that preserve the length of vectors, acts on X by multiplying it with elements M from SU(2). Since M is unitary, this multiplication ensures that the length of X remains unchanged. This means SU(2) performs rotations on the vector X. Conversely, any transformation that preserves the properties of trace-zero Hermitian matrices must be unitary, proving that every rotation in three-dimensional space corresponds to an element in SU(2). However, each rotation in SO(3) is associated with two different elements in SU(2): M and -M. This relationship confirms that SU(2) is a double cover of SO(3).
Furthermore, SU(2) is simply connected, a property that means there are no "holes" or obstructions in its structure. This can be understood by recognizing that SU(2) is equivalent to the group of unit quaternions, which is a mathematical space similar to a 3-sphere.
A unit quaternion consists of a scalar part and a vector part. The scalar part equals the cosine of half the rotation angle, while the vector part equals the sine of half the rotation angle multiplied by a unit vector along the rotation axis. If a rigid object starts in a specific orientation (represented by a quaternion with a scalar part of +1 and a vector part of zero), after one full rotation (2π radians), the scalar part becomes -1, and the vector part returns to zero. After two full rotations (4π radians), the scalar part returns to +1, and the vector part again becomes zero, completing the cycle.