home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!paladin.american.edu!gatech!usenet.ins.cwru.edu!agate!doc.ic.ac.uk!doc.ic.ac.uk!usenet
- From: sjv@doc.ic.ac.uk
- Newsgroups: sci.logic
- Subject: Re: Product of sites
- Date: 11 Jan 1993 12:58:55 GMT
- Organization: Imperial College
- Lines: 30
- Distribution: world
- Message-ID: <1irqufINNhu8@frigate.doc.ic.ac.uk>
- References: <ARA.93Jan10125554@camelot.ai.mit.edu>
- NNTP-Posting-Host: dse-mac-l1-89.doc.ic.ac.uk
-
- In article <ARA.93Jan10125554@camelot.ai.mit.edu> ara@zurich.ai.mit.edu (Allan
- Adler) writes:
- >
- >Does the category of sites have products and if so how does one construct
- them?
- >
- >It follows from material I found in Johnstone's book Topos Theory
- >that the category of Grothendieck toposes has products but I don't know
- >if the same is true for sites in general.
- >
-
- You don't say what your morphisms are between sites (nor between toposes for
- that matter).
-
- One sensible approach is to define the morphisms between sites so that they
- correspond to geometric morphisms between the corresponding sheaf toposes. Then
- the categories of sites and of Grothendieck toposes are equivalent, so of
- course the category of sites has (finite) products.
-
- Details will depend on exactly how you define site - e.g. is it an arbitrary
- category with Grothendieck topology, or a category with finite limits? But what
- it boils down to is that the site is some kind of presentation by generators
- and relations for the topos, so a morphism is an interpretation of one site in
- the sheaf topos of the other.
-
- If you are considering sites based on arbitrary categories, then I guess the
- product is going to be a "disjoint union" of sites. (Remember that geometric
- morphisms in a sense go backwards, so limits look like colimits.)
-
- Steve Vickers.
-