On automorphisms of categories with applications to universal algebraic geometry (Q6089423)

Let \(\mathcal{V}\)\ be a variety of algebras of some type \(\Omega\). B. Plotkin suggested a new method of studying properties of \(\mathcal{V}\)-algebras using some analogy to classical algebraic geometry and called the corresponding theory the universal algebraic geometry [\textit{G. Mashevitzky} et al., Electron. Res. Announc. Am. Math. Soc. 8, 1--10 (2002; Zbl 0996.08008); \textit{B. Plotkin}, Isr. J. Math. 96, Part B, 511--522 (1996; Zbl 0921.08004); Sib. Adv. Math. 7, No. 2, 64--97 (1997; Zbl 0936.08007); in: Matematicheskaya logika i algebra. Sbornik statej. K 100-letiyu so dnya rozhdeniya akademika Petra Sergeevicha Novikova. Moskva: Maik Nauka/Interperiodika. 165--196 (2003; Zbl 1062.08012); \textit{B. Plotkin} and \textit{E. Plotkin}, Demonstr. Math. 48, No. 4, 578--619 (2015; Zbl 1344.08003); \textit{E. Aladova} et al., J. Math. Sci., New York 214, No. 6, 742--754 (2016; Zbl 1344.08006); translation from Fundam. Prikl. Mat. 19, No. 3, 5--22 (2014); \textit{A. Tsurkov}, Int. J. Algebra Comput. 17, No. 5--6, 1263--1271 (2007; Zbl 1162.08002); \textit{A. Tsurkov}, Appl. Categ. Struct. 24, No. 3, 209--240 (2016; Zbl 1359.08003)]. The method is to find all terms in the language of a given variety, which determine such \(\Omega\)-algebras that are isomorphic to a given \(\Theta^{0}\left( \mathcal{V}\right) \)-algebra and have the same underlying set with it. However this method can be applied only to automorphisms which take all objects to isomorphic ones. This paper, being an improved and augmented version of the author's preprint [``On automorphisms of categories'', Preprint, \url{arXiv:math.GM3589936}], aims to fill up this gap, using a new type of automorphisms called the \textit{quasi-inner automorphisms} (Definition 2.1), which is enough to characterize arbitrary automorphisms. Further reasoning essentially follows the ideas of [\textit{B. Plotkin}, in: Matematicheskaya logika i algebra. Sbornik statej. K 100-letiyu so dnya rozhdeniya akademika Petra Sergeevicha Novikova. Moskva: Maik Nauka/Interperiodika. 165--196 (2003; Zbl 1062.08012); \textit{B. Plotkin} and \textit{G. Zhitomirski}, J. Algebra 306, No. 2, 344--367 (2006; Zbl 1105.08006)], but a new more successful method is proposed instead of the method described above.This method reduces the problem to the case when the underlying set \(\left\vert A\right\vert \)\ of an algebra \(A\)\ is a subset of the underlying set of the algebra \(\Phi\left( A\right) \), and every endomorphism \(\mu\)\ of the algebra \(A\)\ is the restriction of the endomorphism \(\Phi\left( \mu\right) \)\ of the algebra \(\Phi\left( A\right) \)\ to the set \(\left\vert A\right\vert \)\ and every restriction to \(\left\vert A\right\vert \)\ of an endomorpism \(\upsilon \)\ of \(\Phi\left( A\right) \)\ is an endomorphism \(\mu\)\ of \(A\)\ such that \(\Phi\left( \mu\right) =\upsilon\). The conditions provide an opportunity to describe the action of the endormorphism \(\Phi\). The main results are formulated in two theorems. Theorem 3.1 claims that every automorphism of a category \(\mathcal{C}\)\ supplied with a forgetful functor \[ \boldsymbol{Q}:\mathcal{C}\rightarrow\mathsf{Set} \] \ abiding bt two additional plain conditions is potentially quasi-inner. Theorem 5.3 states that every automorphism \(\Phi\)\ of the category \(\Theta ^{0}\left( \mathcal{V}\right) \)\ for an arbitrary variety \(\mathcal{V}\)\ is the product of two functors \(\Phi=\Gamma\circ\Psi\). The first of them is an inner isomorphism \[ \Psi:\Theta^{0}\left( \mathcal{V}\right) \rightarrow\mathcal{D} \] where \(\mathcal{D}\)\ is a special full subcategory of \(\Theta^{0}\left( \mathcal{V}\right) \). The second functor \[ \Gamma:\mathcal{D}\rightarrow\Theta^{0}\left( \mathcal{V}\right) \] is a so-called extension functor, that is to say, \(\left\vert A\right\vert \subseteq\left\vert \Gamma\left( A\right) \right\vert \) for every \(\mathcal{C}\)-algebra \(A\)\ and \(\mu\subseteq\left\vert \Gamma\left( \mu\right) \right\vert \) for every \(\mathcal{C}\)-morphism \(\mu\). The latter part of \S 5 contains two examples. The first of them shows that the advantage of this new method in a known situation (the variety of all semigroups), while the second one presents a new result for the variety \(\mathcal{M}\)\ of all modules over arbitrary ring \(R\)\ with unit. The last \S 6 is devoted to applications to universal algebraic geometry.