Generational turnover means hard lessons have to be learned again from scratch

The argument

A second argument that history repeats itself turns on who is in the room. Proponents point out that the people who paid for a disaster are not the people making decisions a few decades later. Experience of a currency collapse, an occupation, an epidemic or a war is held in the bodies of those who lived through it and governs their behaviour for the rest of their lives; it does not transfer to their grandchildren, who hold the same events as information rather than as caution. Proponents argue that the institutions built in the aftermath of a catastrophe are therefore dismantled, on average, by the first generation with no memory of what they were for. The mechanism is reinforced by the rewards available. Safeguards are costly and their benefit is invisible while they are working, so there is always a case for relaxing them, and whoever relaxes them is usually rewarded before the consequences arrive. Proponents argue that this produces a predictable cycle: crisis, protective rules, a long quiet period in which the rules look like needless friction, their removal, and then the recurrence that justified them. On this account the interval between repetitions is set roughly by how long a generation remembers. Proponents regard this as the strongest form of the claim because it specifies a mechanism and a timescale rather than asserting an eternal pattern. They do not say the repetition is exact, only that the conditions for it are rebuilt as predictably as memory fades, and that the same arguments against safeguards reappear almost word for word each time.

Premises

Counter-arguments

Critics accept the memory mechanism and deny that it establishes repetition. Institutional knowledge, they argue, is retained in ways personal memory is not, through written rules, professional training, statistical records and the accumulated design of systems, which is why many protections built after past disasters are still standing. They add that the cycle described predicts far more recurrences than actually occur: most eroded safeguards are not followed by the catastrophe they were built against, which suggests other conditions do most of the work. On this account the pattern is a tendency at best rather than a repetition.

Rejecting the premises

[Rejecting P1] Institutional memory is not the same as personal memory and is retained in written rules, training, records and the design of systems, which outlast the generation that built them. [Rejecting P2] Safeguards are often retained precisely because their rationale is documented, and those who relax them are frequently blamed rather than rewarded once the record is public. [Rejecting P3] The cycle predicts far more recurrences than actually happen, since most relaxed safeguards are not followed by the catastrophe they were built against. [Rejecting C] A mechanism that makes recurrence more likely in some conditions supports a claim about risk, not the claim that history repeats.