Identify the reflective subcategory with its full replete image, writing for the reflection unit. An object is fixed by exactly when precomposition with every gives a bijection . One implication is the reflection adjunction; for the other, take to obtain a retraction of , and uniqueness of maps into the fixed object makes it a two-sided inverse.
An ambient finite limit of fixed objects has this same property: its hom-sets are limits of the hom-sets of the diagram, and a limit of these bijections is a bijection. Thus the ambient terminal object, products and equalizers are fixed and retain their universal properties in the full subcategory. The reflective subcategory is Cartesian, with finite limits computed in the ambient topos.
Let be fixed objects and form the ambient exponential object . For every , the exponential adjunction identifies maps with maps . Since is left exact and ,
and its unit corresponds to . Because is fixed, every map therefore extends uniquely to . Transposing back gives
The fixed-object criterion from part (i) shows that already belongs to . The ambient evaluation and transpose universal property restrict to the full subcategory, proving the subcategory is cartesian closed. This constructs exponentials in a left-exact reflective subcategory, rather than merely reflecting an exponential and assuming its universal property survives.
For a subobject , define the closure operation induced by a left-exact reflector by
where the pullback uses and the monomorphism . Naturality of the unit factors through , proving extensivity, and functoriality of proves monotonicity. Applying the left-exact to the defining pullback gives , since is invertible. Repeating the construction therefore changes nothing, proving idempotence.
For , naturality of and preservation of pullbacks identify the closure of with , which is also . Hence
The operation is thus universal: it is stable under every pullback. The same construction preserves finite intersections because is left exact.
Every characteristic map expresses its subobject as . Universality of the closure operation induced by a left-exact reflector gives
By definition, classifies , so the right-hand subobject has characteristic map . The original and its closure agree exactly when their unique characteristic maps agree. Thus
Extensivity, idempotence, preservation of intersections and closure of the whole subobject also give the Lawvere-Tierney topology laws for .
Let be the equalizer of and the identity. By part (iv), maps correspond exactly to closed subobjects of . Closed subobjects of in turn correspond to subobjects of in : send to , and in the reverse direction pull a fixed subobject back along . Left exactness shows that these operations are inverse; the pulled-back object has reflection , making it closed by the defining formula.
Consequently
Replacing by and using invertible shows that maps into extend uniquely across every unit. The criterion of part (i) therefore proves itself is fixed, so belongs to . Truth factors through it because . For fixed , the displayed bijection becomes
With the factored truth arrow this is precisely the subobject classifier universal property. Thus the closed-subobject classifier supplies the classifier in the reflective subcategory.

Articles by others on the same topic (0)

There are currently no matching articles.