0
$\begingroup$

In section 12 of the book Surfaces in classical geometries: A treatment by moving frames by Gary R. Jensen, Emilio Musso and Lorenzo Nicolodi (see preview here), Möbius geometry is described.

They introduce the generalized Minkowski space of signature $(4,1)$ as follows:

Let $R^{4,1}$ denote $R^5$ with a Lorentzian inner product. Let $\epsilon_0,\dots,\epsilon_4$ denote the standard orthonormal basis of $R^{4,1}$ given by the standard orthonormal basis $\epsilon_0,\dots,\epsilon_3$ of the Euclidean space $R^4$ and with $\langle \epsilon_4, \epsilon_4\rangle=-1$. The Lorentzian inner products $\langle \epsilon_a, \epsilon_b\rangle$, for $a,b=0,\dots,4$, are the entries of the matrix $$\begin{pmatrix} I_4 & 0 \\ 0 & -1 \end{pmatrix}.$$ Write elements of $R^{4,1}=R^4\oplus R\epsilon_4$ as $x + t\epsilon_4$, where $x\in R^4$ and $t\in R$. The Lorentzian inner product is then $$\langle x+s\epsilon_4,y+t\epsilon_4\rangle=x\cdot y-st.$$

My problem is that I would expect vectors in $R^{4,1}$ to have dimension $5$, just like vectors in the Minkowski space have dimension $4$. They actually start stating that $R^{4,1}$ is $R^5$ with a Lorentzian inner product. This could suggest that it is just a typo (two actually), but later on in the chapter they are consistent on this matter: vectors in $R^{4,1}$ have $4$ components. For instance, they consider a mapping which is the sum of a vector in $\mathbb{S}^3$ with the vector $\epsilon_4$. Therefore I assume there's no typo and it's just my poor understanding...

Can anyone help understanding this construction and how it works?

$\endgroup$
  • $\begingroup$ I'm confused. Where exactly do you see a typo or the statement that elements of $R^{4,1}$ have $4$ components? $\endgroup$ – freakish Oct 13 '18 at 13:33
  • $\begingroup$ Note the four vectors $\epsilon_0,\epsilon_1,\epsilon_2,\epsilon_3$ give the +ve subspace and $\epsilon_4$ is -ve subspace of the inner product. The slightly nonstandard way of indexing the +ve subspace probably caused your confusion. (By the way, the usual definition of signature is not going to give (5,1) but (4,1)). $\endgroup$ – user10354138 Oct 13 '18 at 13:45
  • $\begingroup$ @freakish They write vectors of $R^{4,1}$ as $x+t\epsilon_4$, where $x\in R^4$. This plus the fact that the $\epsilon_i$ for $i=0,\dots,3$ are the standard orthonormal basis of $R^4$ makes me think that $x+t\epsilon_4$ is a sum of $4$-vectors... What am I getting wrong? $\endgroup$ – Edu Oct 13 '18 at 14:14
  • $\begingroup$ @user10354138 That's clear but again, the say that $\epsilon_0,\dots,\epsilon_3$ is the standard orthonormal basis of $R^{4}$... how is that they don't have $4$ components then? $\endgroup$ – Edu Oct 13 '18 at 14:16
  • 1
    $\begingroup$ @Edu This is not a sum of $4$-vectors. First of all note that $e_4\in R^5$. What they say is that because $R^{4,1}=R^4\oplus R$ then every 5-vector can be written as $x+te_4$ where $x\in R^4\oplus 0$ and $e_4\in 0^4\oplus R$ in a unique way. When they write $x\in R^4$ what they actually mean is that $x$ is in a $4$-dimensional subspace of $R^5$, i.e. formally $x\in R^4\oplus 0$ meaning $x=(x_1,x_2,x_3,x_4, 0)$. In other words they say that any 5-vector can be written as $(x_1,x_2,x_3,x_4,0)+t(0,0,0,0,1)$ which is quite obvious. $\endgroup$ – freakish Oct 13 '18 at 14:27
0
$\begingroup$

So when they write

Write elements of $R^{4,1}=R^4\oplus Re_4$ as $x+te_4$, where $x\in R^4$ and $t\in R$.

what they actually mean is that $x$ is in the $4$-dimensional subspace of $R^{4,1}$ spanned by $\{e_1,e_2,e_3,e_4\}$. It doesn't mean that $x\in R^4$ literally, $x$ is still an element of a $5$-dimensional space but one of the coordinates is zerod.

For example if we take the standard basis $e_i=(0,\ldots,0,1,0,\ldots, 0)$ with $1$ at $i$-th posisition then the statement can be rewritten as

Write elements of $R^{4,1}$ as $(x_0,x_1,x_2,x_3,0)+t(0,0,0,0,1)$.

More generally the precise statement should be

Write elements of $R^{4,1}$ as $x+te_4$ where $x=x_0e_0+x_1e_1+x_2e_2+x_3e_3$ for $x_0,x_1,x_2,x_3,t\in R$.

Here $R^4$ means $span(e_0,e_1,e_2,e_3)$. It is a $4$-dimensional subspace of $R^5$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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