Markov Chain
A process whose future depends only on the present state
A Markov chain is a stochastic process with the Markov property: the conditional distribution of the next state depends only on the current state, not on the full history that led there. Named for Andrey Markov (1856–1922); in information theory the same idea appears whenever a processing step is sufficient for what comes after — the past beyond that step is irrelevant to the output.
In Genesis, a pipeline of artifacts is modelled as a Markov chain when each step consumes the previous artifact as its complete input: ground truth X → intermediate Y → derived Z, with Z depending on X only through Y. That structure is what makes the data processing inequality apply: I(X ; Z) ≤ I(X ; Y) — mutual information with the source cannot increase along the chain.
When a step secretly re-reads earlier inputs, merges uncited side channels, or joins streams without documented ordering, the Markov assumption breaks and the inequality may not bound trust composition — which is why causal chains and cross-stream consistency must be explicit in L1–L3 artifacts.
Corpus stance
A1 — adopted on this site: Structural condition under which Shannon’s processing limits apply to the atomicity ladder; see What Information Theory Says.