Vladimir Voevodsky writes:
> The slice c=
ategory over P is equivalent to presheaves on the category of elements of P=
. What is your definition of the category of elements of $P$? Objects are p=
airs $(X:C,p:P(X))$, and morphisms from $(X,p)$ to $(X=E2=80=99,p=E2=80=99)=
$ are=E2=80=A6?
...morphisms $f : C(X',X)$ suc=
h that $Pf(p) =3D p'$.
The slice category of t=
he category of functors $A -> Set$ over a functor $F$ is equivalent to f=
unctors $el(F) -> Set$, right? =C2=A0Now if you define presheaves as fun=
ctors $C^{op} -> Set$, you get that presheaves on $C$ over a presheaf $P=
$ are equivalent to *functors* $el(P) -> Set$, not presheaves on $el(P)$=
.
That's why I was trying to be clever and use=
that unnatural definition of presheaves, but, as you pointed out, that doe=
sn't work either, because the resulting category of presheaves is the o=
pposite of what we want.
Paolo
------=_Part_260_639369971.1487948801043--
------=_Part_259_552548430.1487948801043--