Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Someone has claimed that he has constructed a quaternion representation of the one dimensional (along the x axis) Lorentz Boost.

His quaternion Lorentz Boost is $v'=hvh^*+ 1/2( [hhv]^*-[h^*h^*v^*]^*)$ where h is (sinh(x),cosh(x),0,0). He derived this odd transform by substituting the hyperbolic sine and cosine for the sine and cosine in the usual unit quaternion rotation $v'=hvh^*$ and then subtracting out unwanted factors. You can see the short "proof" here:

I have argued that, whereas the quaternion rotations form a group, his newly devised transform probably does not. Two transformations of the form $v'=hvh^*+ 1/2( [hhv]^*-[h^*h^*v^*]^*)$ almost certainly do not make another transformation of the form $v'=fvf^*+ 1/2( [ffv]^*-[f^*f^*v^*]^*)$ He has not attempted to prove that and he won't because he thinks it is unnecessary.

He has responded that, even if I'm correct, he still has created a Lorentz Boost despite the fact that he has not created a group. I've argued that that it is essential that two Lorentz Boosts make another Lorentz Boost. Without that group structure there is no Boost. Is this correct?

Can you have a Lorentz Boost along the x axis without having the group structure Boost+Boost=Boost? (for the case of Boosts along the x axis. I know that, for boosts in different directions, Boost+Boost=Rotation)

share|cite|improve this question

Boost along the $x$ direction. The result is $X' = \Lambda(\phi) X$, where $$X = \left( \begin{array}{c} c t \\ x\end{array}\right) \hspace{5ex} \textrm{and} \hspace{5ex} \Lambda(\phi) = \left( \begin{array}{cc} \cosh \phi & -\sinh\phi \\ -\sinh\phi & \cosh \phi \end{array}\right),$$ and where $\phi$ is the rapidity.

Boost again along the $x$ direction, $X'' = \Lambda(\phi')X' = \Lambda(\phi')\Lambda(\phi) X$. Using the properties of the hyperbolic trigonometric functions, we can see that $X'' = \Lambda(\phi'') X$, where $\phi'' = \phi + \phi'$. (Do the matrix multiplication, $\Lambda(\phi')\Lambda(\phi)$.) Thus, the composition of boosts along the $x$ direction is another boost along the $x$ direction. The rest of the group axioms are also satisfied. Thus, the collection of boosts along the $x$ direction form a group.

The group structure is independent of the representation. If the "representation" is not closed under composition, it is definitely not a representation of the group.

Lorentz transformations can be represented by quaternions. See here for example.

share|cite|improve this answer

This is really just a matter of deciding if his idea to call this a "Lorentz boost" is justified or not. It's a choice of terminology, not a fixed mathematical idea that needs a mathematician to decide whether it is right or wrong.

For regular Minkowski space, the composition of two boosts need not be a pure boost: the composition generally incorporates a spatial rotation. So I don't see any good reason that boosts should be required to form a group (although all of the Lorentz transformations form a group.)

Edit: After rereading I see that you probably know that, but are just questioning whether or not boosts composing to a nonboost would occur in a 1-dimensional setting. Unless his definition is a special case or version of higher dimensional cases, this could be debatable. Be sure to post in the physics forum too!

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.