You can get a vector field from a pair of spinor fields with $A_\mu(x)=\psi(x) \gamma_\mu \overline{\psi}(x)$. Using this fact could you define a space-time vector in terms of Grasman numbers?
Say you had 16 Grassmann numbers and their conjugates. If you defined a coordinate:
$$ x_\mu \equiv \Theta ^\alpha \Gamma^{\alpha\beta}_\mu \overline{\Theta}^\beta\tag{1}$$
And then fields depended on this variable. $\phi(x)$. Could you get a consistent field theory? You would still have $[x_\mu,x_\nu]=0$ but the only difference I can see is that no function of $x$ could have powers greater than 16.
How would this effect physics? It would be the same as normal physics except for the strange rule:
$$|x|^{17}=0\tag{2}$$
Or in other words the coefficients of the fields $\phi(x)$ would be zero after the 17th term. Would this be disastrous?! You wouldn't be able to have functions like $e^{-x^2}$ as they would terminate after the 17th term. Unless you might define a coordinate $y$ as:
$$ x_\mu \equiv e^{-|y|^2}y_\mu\tag{3}$$
and take the $y$ as the space-time coordinates.
But on the other hand there are stranger algebras such as non-commutative algebra.