Dualisability. Unary algebras and beyond (Q2573123)

Perhaps the earliest example of the use of a natural duality to solve a hitherto intractable problem is Galois theory. Later examples are Pontryagin's duality for abelian groups, Stone's duality for Boolean algebras and \textit{H. A. Priestley}'s duality for distributive lattices. It is perhaps the last example [Bull. Lond. Math. Soc. 2, 186--190 (1970; Zbl 0201.01802); Proc. Lond. Math. Soc. (3) 24, 507--530 (1972; Zbl 0323.06011)] which triggered the treasure hunt for dualisable algebras which has occurred in the past two decades: the property of dualisability has become of great interest in its own right rather than just being a tool for solving other problems. This fascinating book shows that even in the realm of unary algebras, where one might expect the situation to be virtually trivial (especially if one were an ex group theorist like the reviewer) almost all pathological behaviour occurs already. A duality for an algebra \(\underline{\mathbf M}\) is constructed in terms of the quasivariety \(\mathcal A\) it generates. What is sought is a topological quasivariety \(\mathcal X\) and a pair of contravariant functors between \(\mathcal A\) and \(\mathcal X\) satisfying appropriate conditions. We obtain a \textit{full duality} if the categories \(\mathcal A\) and \(\mathcal X\) are dually equivalent and a \textit{strong duality} under an apparently slightly stronger, but more amenable, condition. Chapters 1 and 2 lay the foundations and provide many useful examples. The heart of the book lies in Chapter 3, where the dualisability problem for three-element unary algebras is solved completely, but certainly non-trivially. Even with such small algebras one can obtain examples, such as a chain of clones, where the corresponding algebras are alternately dualisable and non-dualisable. It is shown also that all finite unars (unary algebras with one fundamental operation) are dualisable. Chapter 4 builds on the results of Chapter 3, and determines which of the dualisable three-element unary algebras are also fully or strongly dualisable. It turns out that in this case these two concepts are equivalent. Chapter 5 examines the question as whether dualisability is preserved under various algebraic operations. The answer, sadly, is ``not often''; indeed, an example is given to show that even the product of two dualisable algebras can be non-dualisable. In Chapter 6 more examples of pathological behaviour are constructed and Chapter 7 looks at algebras which are \textit{inherently non-dualisable}, that is, cannot be embedded in a dualisable algebra. Two important general results on strong dualisability which were used in the body of the text are proved in the appendix. The list of references is thorough and the index excellent.