arXiv2015
Modified gravity which was constructed by torsion scalar $T$, namely $f(T)$ doesn't respect Lorentz symmetry. As an attempt to make a new torsion based modified gravity with Lorentz invarianve, recently $f(T,\mathcal{B})$ introduced where $B=2\nabla_μT^μ$ \citep{Bahamonde:2015zma}. We would argue, even when all is constructed and done in a self-consistent form, if you handle them properly,we observe that there is no Lorentz invariant teleparallel equivalent of $f(R)$ gravity. All we found is that the $f(R)$ gravity in which $R$ must be computed in Weitzenböck spacetime, using Weitzenböck's connection, nor Levi-Civita connections is the only possible Lorentz invariant type of modified gravity. Consequently, $f(T)$ gravity can not obey Lorentz symmetry not only in its orthodoxica form but even in this new framework $f(T,\mathcal{B})$.