It's false: consider the following as a quadratic form on $\mathbb{R}^2$: $$\Phi((x_1,x_2))=x_1^2-x_2^2$$
then $(1,0)$ and $(0,1)$ are both anisotropic but $\Phi((1,0)+(0,1))=0$
It's false: consider the following as a quadratic form on $\mathbb{R}^2$: $$\Phi((x_1,x_2))=x_1^2-x_2^2$$
then $(1,0)$ and $(0,1)$ are both anisotropic but $\Phi((1,0)+(0,1))=0$