I'm talking with Jarik about polymorphisms.

Me: We can always assume that they are idempotent. Bulatov Jeavons and Krohkin do this.
Jarik: Then your essentially sub-unary automatically gives projective.
Me: Yeah.
Jarik: Then we have it.
Me: Ah, no, not quite. We can't assume the set is conservative, so it doesn't have to be projective.
Jarik: Fascisms!