The Epigenetic Tempo

Evolutionary argument usually runs on two clocks. The slow one measures allele frequency change across generations, and it is slow enough that anything happening inside a human lifetime is, by definition, not evolution. The fast one measures cultural change, and it is fast enough that we treat it as a separate subject with its own departments.

The gap between them is where most of the confusion about human change lives. A cultural condition arrives in a decade; a genetic response takes centuries; therefore, the argument goes, culture cannot be doing biological work on any timescale we can observe.

There is a third clock. Methylation patterns, chromatin accessibility, histone modification, and small-RNA populations respond to conditions within a lifetime — sometimes within weeks — and some fraction of that response is transmitted to the next generation before selection has taken any position on it. It is not Lamarckian inheritance and it is not allele frequency change. It is a distinct tempo, and it is the one that makes the timescales in this arc plausible rather than fanciful.

The regulatory layer decides which parts of a genome are readable, in which tissues, under which conditions. Two people with nearly identical sequence can differ substantially in stress physiology, immune threshold, metabolic set point, and circadian entrainment because their regulatory settings were laid down inside different environments.

Those environments are largely cultural. Shift work is a cultural fact with a circadian consequence. Chronic threat signalling — from precarity, from conflict, from an information diet engineered for arousal — is a cultural fact with a cortisol consequence. Diet composition, light exposure after dusk, interruption frequency, air chemistry: each is a cultural arrangement that reaches the regulatory layer without asking permission from the sequence.

This is why the arc treats culture as a causal input rather than a commentary. Culture does not need to change allele frequencies to change bodies. It has a faster route, and that route runs through regulation.

The connection between this chapter and the last is structural, not poetic, and it is worth stating precisely so that it cannot be inflated later.

Tunneling supplies variation the genome did not request: a floor of change that arises from physics and is held down by enzymatic work. Regulation supplies responsiveness the genome did not have to evolve: a capacity to reconfigure expression inside a lifetime in answer to conditions.

Both move faster than the standard picture, in which the genome proposes and the environment disposes. Together they describe an organism that is more porous to its circumstances at both ends — at the level of the base pair and at the level of the regulatory state — than the Modern Synthesis needed it to be.

STATUS — The mechanisms are established individually. The claim that they combine into a materially faster route from environment to heritable change is this arc's model, not settled science.

Transgenerational epigenetic inheritance is far better established in plants and nematodes than in humans. In plants, marks can persist for many generations. In mammals the germline resets much of the pattern, twice, and the resets are thorough.

The strongest human claims rest on cohorts small enough to argue about — famine studies, wartime exposures, a handful of natural experiments where the exposure was sharp and the records unusually good. The effects reported are real but modest, the confounds are severe, and the field has a history of over-reading.

This chapter is written to be constrained by that literature rather than to outrun it. What it needs is not dramatic multigenerational inheritance. It needs only what is defensible: that regulatory state responds to environment within a lifetime, that some of that state is transmitted at least to the following generation, and that the transmitted fraction is not zero.

If it turns out to be effectively zero in humans — if the germline resets are complete enough that nothing cultural crosses the generational boundary except by ordinary learning — then this arc loses its fast intermediate clock and has to fall back on demographic sorting alone. That would not destroy the argument, but it would slow it by an order of magnitude, and the chapters on present-day sorting would have to be rewritten as chapters about the next several centuries.

There is a feature of the epigenetic layer that cuts against every determinist reading of this arc, and it should be stated here rather than saved for the end: regulation is reversible in ways sequence is not.

A methylation pattern laid down by a decade of night work is not a permanent modification of a lineage. It is a setting, and settings can be changed by changing the conditions that set them. Whatever differentiation the fast layer produces between culturally separated populations is provisional in a way that genetic differentiation is not.

This is the difference between a description of pressure and a prophecy of outcome. The pressure is measurable. The outcome depends on what happens next, including what people decide to do — which is why the later chapters of this arc are about legibility and method rather than about fate.