Lemma 8.3.3. In the situation of the preceding definition, the pair \[ \bigl (\LMod _R(C)_{\geq 0},\LMod _R(C)_{\leq 0}\bigr ) \] defines a t-structure on \(\LMod _R(C)\). Its truncation functors are computed on underlying objects.

Proof. The adjunction between the symmetric monoidal inclusion \(C_{\geq 0}\hookrightarrow C\) and its lax symmetric monoidal right adjoint \(\tau _{\geq 0}\) induces, after passing to algebras and modules and taking the fiber over \(R\), an adjunction \[ \LMod _R(C_{\geq 0})\rightleftarrows \LMod _R(C). \] Its right adjoint is computed on underlying objects by \(\tau _{\geq 0}\). Finite limits and colimits of modules are computed on underlying objects by Proposition 17.1.16. Since the forgetful functor detects isomorphisms, \(\LMod _R(C)\) is stable. Closure under the required shifts follows from the corresponding property in \(C\), so it remains to check orthogonality and decomposition. If \(X\) is connective and \(Y\) is coconnective, adjunction gives \[ \Hom _{\LMod _R(C)}(X,Y[-1]) \simeq \Hom _{\LMod _R(C_{\geq 0})}(X,\tau _{\geq 0}(Y[-1])) \simeq *. \] Finally, for an arbitrary \(R\)-module \(X\), the counit \(\tau _{\geq 0}X\to X\) in modules has connective source, and its cofiber belongs to \(\LMod _R(C)_{\leq -1}\) because this is true after forgetting to \(C\). This gives the required decomposition. □