The \(-_{+}\) and \(-^{+}\) constructions for biset functors (Q1714865)

The first author introduced a canonical Brauer induction formula for the ordinary character ring of a finite group \(G\) in [Astérisque 181--182, 31--59 (1990; Zbl 0718.20005)]. Then he used a systematic axiomatization of the problem to obtain a canonical induction formula for every Mackey functor in [J. Algebra 206, No. 1, 293--343 (1998; Zbl 0913.20001)]. Mackey functors where at the time the main functorial framework for the study of the induction theories. However, it has two unsatisfactory aspects. The first is the restriction to subgroups of a \emph{fixed finite group} and the second is the impossibility to deal with relations between a group and its quotients via \emph{inflation} and \emph{deflation}. The theory of biset functors introduced by \textit{S. Bouc} [J. Algebra 183, No. 3, 664--736 (1996; Zbl 0858.19001)] solves both of these problems by giving the freedom of choosing categories where the morphisms are constructed from the `elementary bisets' operations of restriction, induction, inflation, deflation and isomorphism. The main result of this article is the extension of the two \(+\)-constructions of the first author to this setting. This is the first step for the definition and construction of canonical induction formulae for representation rings viewed as biset functors. When \(\mathcal{D}\) is a non-full subcategory of a category of bisets, the authors define two categories \(\mathcal{D}_{+}\) and \(\mathcal{D}^{+}\) having the same objects and such that \(\mathcal{D} \subseteq \mathcal{D}_{+}\subseteq \mathcal{D}^{+}\). The category \(\mathcal{D}_{+}\) is obtained by adding all the possible inductions to \(\mathcal{D}\) and \(\mathcal{D}^{+}\) is obtained by adding all possible inductions and restrictions to \(\mathcal{D}\). The other elementary operations are not changed. The \(-_+\) construction is then a functor from \(\operatorname{Fun}_{R}(\mathcal{D},R)\) to \(\operatorname{Fun}_R(\mathcal{D}_{+},R)\) and the \(-^+\) construction is a functor from \(\operatorname{Fun}_{R}(\mathcal{D},R)\) to \(\operatorname{Fun}_R(\mathcal{D}^{+},R)\). The authors have carefully worked out what conditions to impose on the category \(\mathcal{D}\) for this two results to hold and it is difficult to imagine a concrete situation where we cannot apply the \(-_+\) construction. The other requires more constrains on the category \(\mathcal{D}\) to hold. The two constructions are rather subtle and this is enlightened in the last section of the article where the authors show that in general \(-_+\) and \(-^+\) are \emph{not} exactly the left and the right adjoints to the forgetful functor!