Not signed in (Sign In)

Not signed in

Want to take part in these discussions? Sign in if you have an account, or apply for one below

  • Sign in using OpenID

Discussion Tag Cloud

Vanilla 1.1.10 is a product of Lussumo. More Information: Documentation, Community Support.

Welcome to nForum
If you want to take part in these discussions either sign in now (if you have an account), apply for one now (if you don't).
    • CommentRowNumber1.
    • CommentAuthorDavidRoberts
    • CommentTimeAug 20th 2010

    Cross-posted from MO. Feel free to put any answers there to pointy-licious goodness


    I’m working out of Sheaves in geometry and logic, for reference.

    There is a characterisation of flat functors A:CSetA:C \to Set as those such that the Grothendieck construction CA\int_C A is a filtering category. There are more general versions of this result, in which SetSet is replaced by a more general topos. One should also be able to characterise those discrete opfibrations that arise from flat functors (up to iso/equiv?). How about if we replace CC by an internal category, in a topos EE say? Then functors out of CC are replaced by discrete opfibrations over CC in EE.

    My question is this:

    What sort of thing should be considered as the analogue of a flat functor in the internal setting?

    • CommentRowNumber2.
    • CommentAuthorTobyBartels
    • CommentTimeAug 20th 2010
    • CommentRowNumber3.
    • CommentAuthorFinnLawler
    • CommentTimeAug 20th 2010

    I’ve mumbled something in response over at MO.

    • CommentRowNumber4.
    • CommentAuthorDavidRoberts
    • CommentTimeAug 21st 2010

    Thanks, Finn. I’ve accepted your answer.

    • CommentRowNumber5.
    • CommentAuthorFinnLawler
    • CommentTimeAug 21st 2010

    Cool, hope it helped.

    • CommentRowNumber6.
    • CommentAuthorDavidRoberts
    • CommentTimeAug 21st 2010

    It did, but google books doesn’t give me any joy with either of Johnstone’s books, nor the handbook of categorical algebra.

    Could you pretty please copy out the axioms?

    • CommentRowNumber7.
    • CommentAuthorTobyBartels
    • CommentTimeAug 21st 2010

    From Elephant B2.6.1 (slightly shortened):

    We begin by defining a number of objects associated with an internal category CC.

    • PP, the object of pairs of morphisms with common codomain, is defined as the pullback C 1× 0,0C 1C_1 \times_{0,0} C_1.
    • QQ, the object of pairs with common domain, is C 1× 1,1C 1C_1 \times_{1,1} C_1.
    • RR, the object of parallel pairs, is the intersection of PP and QQ, or equivalently the pullback C 1× (0,1),(0,1)C 1C_1 \times_{(0,1),(0,1)} C_1.
    • SS, the object of commutative squares, is C 2× 1,1C 2C_2 \times_{1,1} C_2.
    • TT, the object of diagrams of type hgf\bullet \overset{g}\underset{h}\rightrightarrows \bullet \overset{f}\rightarrow \bullet with fg=fhf g = f h, is C 2× (0,1),(0,1)C 2C_2 \times_{(0,1),(0,1)} C_2.

    These are all defined with diagrams which I hope you can write down for yourself, perhaps by my subscript notation. The pulled back maps in these pullbacks are all given by lowercase letters that match the uppercase letter that names the pullback itself. For example, p 1,p 2:PC 1p_1, p_2\colon P \to C_1, d 0p 1=d 0p 2d_0 p_1 = d_0 p_2, and PP is universal with this structure. The structure maps C 1C 0C_1 \rightrightarrows C_0 themselves are d 0,d 1d_0, d_1 as usual.

    From Elephant B2.6.2:

    An internal category CC in a regular category is filtered if each of the three morphisms C 01C_0 \to 1, (d 1p 1,d 1p 2):PC 0×C 0(d_1 p_1, d_1 p_2)\colon P \to C_0 \times C_0, and (d 2t 1,d 2t 2):TR(d_2 t_1, d_2 t_2)\colon T \to R is a cover.

    Note: in this definition we have adopted the custom of naming morhpisms into a pullback by the names of the morphisms into the product of which the pullback is a subobject.

    • CommentRowNumber8.
    • CommentAuthorFinnLawler
    • CommentTimeAug 21st 2010
    • (edited Aug 21st 2010)

    Er, yes. What Toby said.

    • CommentRowNumber9.
    • CommentAuthorMike Shulman
    • CommentTimeAug 21st 2010

    For one whose interests incline in that direction, it would be a nice exercise to work out those axioms by starting with the usual definition of a flat functor and interpreting it in the internal logic of the topos.

    • CommentRowNumber10.
    • CommentAuthorDavidRoberts
    • CommentTimeAug 21st 2010

    Thanks. Just what I needed. Well, maybe - it is a minor distraction perhaps from the real task - internal saturated anafunctors :)