A replicator is simply something that makes copies of itself. There are hypothetical replicators (e.g., self-catalyzing chemical cycles) that are suspected to be unable to exhibit heritable variation. Variation in any of their constituent molecules would not lead them to produce offspring with those new variant molecules. Copying, such as in DNA replication or in xerox machines, allows any sequence to be remade and then sequence variations to be inherited. This distinction has been used against non-RNA-world hypotheses: without RNA replication (...) systems have no capacity to exhibit heritable variation. However, is copying the only way to have heritable variation? This article suggests that evolution can happen without anything being copied and that, in fact, RNA copying is too complex to arise spontaneously as the first pre-biotic system. There would then be a historical gradation of systems from non-alive to alive in which the capacity to have heritable changes would progressively increase and in which copying, as in template-based RNA-replication, would be a crucial but relatively late event. In addition, this article proposes a way by which RNA replication could have arisen, after systems able to evolve arose, as a crucial innovation and how that innovation spread over existing evolutionary systems and became a major driver of subsequent evolution. (shrink)