Genes are the true units selected by evolution, not organisms

Dawkins's foundational claim, sharpened from his earlier work, is that natural selection acts most fundamentally on genes rather than on the individual organisms that carry them. A gene succeeds or fails across generations based on whether copies of it persist, and organisms are best understood as vehicles genes build to propagate themselves rather than as the primary object selection cares about.

He defines exactly what counts as a gene here: an "active germ-line replicator," any portion of hereditary material whose particular form influences its own probability of being copied into future generations. This deliberately avoids pinning down an exact physical size, since what matters is the replicating unit's influence on its own survival, not its chemical extent.

This reframing changes the question a biologist asks: not "how does this benefit the animal" but "how does this benefit the gene's own propagation," a subtly different question that sometimes yields a different answer.

Takeaway: when a trait seems self-defeating for the organism, ask instead whether it serves the gene's own replication.