If death is claimed to belong to the class of predators, some analyses distinguish actual predators from a hypothetical predator-like instance represented by death. A criticism asks whether this requires two inconsistent real species under one expression.
The issue returns to established convention versus interpretation. Building an imagined member of a class does not abolish the conventional meaning of the class-name.
The hypothetical instance belongs to rhetorical conception, not to zoological classification, so ordinary referents, figurative inclusion, and linguistic assignment must be kept distinct.