CatDat

Implication Details

Assumptions: epi-regularextensiveregular

Conclusions: co-Malceveffective cocongruences

Reason: Suppose we have a coreflexive corelation X+XpErXX+X' \xtwoheadrightarrow{p} E \xtwoheadrightarrow{r} X on XX, where we let XX' be an isomorphic copy of XX for clarity below. Let YY be the equalizer of pi1,pi2:XEp\circ i_1, p\circ i_2 : X \rightrightarrows E. That means that for a generalized element xX(T)x \in X(T), xY(T)x \in Y(T) if and only if p(x)=p(x)p(x) = p(x'). Then by the assumptions pp is a regular epimorphism. By regularity, pp is the coequalizer of its kernel pair, which can be expressed as the equalizer KK of pπ1,pπ2:(X+X)×(X+X)E,p \circ \pi_1,\, p\circ \pi_2 : (X+X') \times (X+X') \rightrightarrows E, where π1,π2\pi_1, \pi_2 are the projections. By distributivity and extensivity, it is sufficient to calculate the equalizer on each "quadrant" of (X+X)×(X+X)(X+X') \times (X+X'), i.e. the four copies of X×XX \times X.
On the X×XX\times X quadrant, for generalized elements x1,x2X(T)x_1, x_2 \in X(T), we have (x1,x2)K(T)(x_1, x_2) \in K(T) if and only if p(x1)=p(x2)p(x_1) = p(x_2). Since pi1p\circ i_1 is a split monomorphism, this is equivalent to x1=x2x_1 = x_2. Thus, the X×XX\times X quadrant of KK is the diagonal of XX. On the X×XX\times X' quadrant, we have (x1,x2)K(T)(x_1, x_2')\in K(T) if and only if p(x1)=p(x2)p(x_1) = p(x_2'). Since r(p(x1))=x1r(p(x_1)) = x_1 and r(p(x2))=x2r(p(x_2')) = x_2, this condition implies x1=x2x_1 = x_2; and then by definition of YY, x1=x2Y(T)x_1 = x_2 \in Y(T). The converse is straightforward. Thus, the X×XX\times X' quadrant of KK is the diagonal of YY. Similarly, the X×XX'\times X quadrant of KK is the diagonal of YY, and the X×XX'\times X' quadrant of KK is the diagonal of XX.
Thus, we get that a morphism h:X+XZh : X+X' \to Z factors through EE if and only if h(x)=h(x)h(x) = h(x) for every generalized element xXx \in X; h(y)=h(y)h(y) = h(y') for every yYy \in Y; h(y)=h(y)h(y') = h(y) for every yYy\in Y; and h(x)=h(x)h(x') = h(x') for every xXx \in X. Clearly this is equivalent to h(y)=h(y)h(y) = h(y') for every yYy\in Y, so in fact EE is the cokernel pair of i1incYi_1 \circ \inc_Y and i2incYi_2 \circ \inc_Y. This means that EE is an effective cocongruence.

Remark: The assumptions are satisfied in particular for every elementary topos. Therefore, every elementary topos has effective cocongruences and is co-Malcev. This special case is Example 2.2.18 in Malcev, protomodular, homological and semi-abelian categories. An alternative proof of this special case is given later in A.5.17.