Search
next up previous
Next: Myles Tierney - Some Up: Category Theory / Théorie Previous: Joan Wick Pelletier -

Walter P. Tholen - Topology based on maps



WALTER P. THOLEN, Department of Mathematics and Statistics, York University, Toronto, Ontario  M3J 1P3, Canada
Topology based on maps


Compared with the large variety of notions on spaces, maps seem to play a minor role in General Topology. We wish to show that it is beneficial to think of fundamental object notions as being induced by particular classes of morphisms. Hence, with any class ${\cal F}$ of morphisms in a category satisfying some weak but essential properties we associate two new classes of morphisms, ${\cal F}^\ast$ and ${\cal F}'$, and two classes of objects, $C{\cal F}$ and $D{\cal F}$, such that, under suitable restriction, both C and D belong to known and fundamental Galois correspondences. For example, for ${\cal F}$ the class of closed continuous maps in the category of topological spaces, ${\cal F}^\ast$and ${\cal F}'$ are the classes of proper and of separated maps, respectively, and $C{\cal F}$ and $D{\cal F}$ are the categories of compact spaces and of Hausdorff spaces, respectively. Other interesting choices for ${\cal F}$ would include the classes of open maps and of dense maps. Among other things we wish to

$\bullet$  explore the interaction of the classes $C{\cal F}$ and $D{\cal F}$ and derive their standard properties which, in the paradigmatic example, gives a surprisingly complete theory of compactness/Hausdorff separation/perfectness

$\bullet$  discuss with the same general methods connectedness/total disconnectness, both at the object and map levels

$\bullet$  show that the required axioms are satisfied for the three basic choices for ${\cal F}$ (as mentioned above for spaces) and are actually available in any category which comes equipped with a (reasonably good) closure operator

$\bullet$  demonstrate that the morphism-object interaction is advantageously reversed by slicing the categories in question, giving us also a safe guideline for the ``right'' fibrewise notions

$\bullet$  apply the theory in particular to Diers' Zariski closure for algebraic sets (with respect to any monad over sets).


next up previous
Next: Myles Tierney - Some Up: Category Theory / Théorie Previous: Joan Wick Pelletier -
eo@camel.math.ca