Replication

Replication is the process by which a system produces copies of itself — whether molecules, cells, organisms, ideas, or computational processes. It is the engine of heredity, evolution, and memetics: without replication, there is no selection; without selection, there is no adaptation.

In biology, replication refers to DNA replication and cellular reproduction. In computation, it describes self-copying programs and the propagation of information across networks. In the context of abiogenesis and dissipative structures, replication marks the threshold where a self-maintaining chemical network becomes capable of producing functional copies — the transition from metabolism to life.

The formal study of replication connects to autocatalysis (self-catalyzing chemical cycles), hypercycles (networks of mutually catalytic replicators), and error threshold theory — the mathematical limit on how much copying error a replicator can tolerate before information degrades faster than selection can preserve it.

The Error Threshold and the Limits of Replication

Replication is not merely a copying process. It is an information-transmission process, and like all information transmission, it is subject to noise. The error threshold is the mathematical limit on how much copying error a replicator can tolerate before its information degrades faster than selection can preserve it. This threshold, first derived by Manfred Eigen in the 1970s, is one of the deepest constraints on the evolution of replication.

Eigen's formulation is stark. Consider a replicator that produces copies of itself with some error rate per base or per symbol. If the error rate is low, the replicator's information — its sequence, its structure, its function — is faithfully transmitted to the next generation, and selection can refine it. If the error rate is high, each copy is so corrupted that the replicator loses its identity before selection can act. The error threshold is the boundary between these two regimes. For a replicator of length L with an error rate ε per symbol, the threshold is approximately ε ≈ 1/L. This means that longer replicators — more complex sequences, more information — require higher fidelity.

The error threshold is the reason that complex life did not evolve until after the evolution of proofreading mechanisms. The RNA world hypothesis proposes that early life used RNA as both information carrier and catalyst, but RNA replication without proofreading is error-prone, and the error threshold limits the length of replicable RNA sequences to a few hundred nucleotides — far too short for the complex machinery of modern cells. The transition to the DNA-protein world, with its dedicated polymerases and repair enzymes, was not a minor improvement. It was the crossing of a threshold that made complex replication possible.

The connection to algorithmic randomness and information theory is direct. A replicator is a compression device: it compresses the information required to build a functional system into a compact sequence. The error threshold is the point at which the compression becomes unstable — where the noise in the channel exceeds the redundancy in the code. In complex systems, this principle generalizes beyond molecular replication. Any system that propagates information — memes, institutions, technologies — faces an error threshold. The collapse of institutional knowledge, the drift of oral traditions, and the degradation of software through successive modifications are all instances of the same principle: information that is not protected by error-correction degrades.

The error threshold is not a biological fact. It is an information-theoretic law. It constrains all replicators, from molecules to civilizations. The evolution of complexity is the evolution of error-correction, and the error threshold is the boundary that separates the simple from the complex.