Anything that has lasted long enough for us to study it almost always looks as though it had been going somewhere. Species toward complexity, languages toward order, institutions toward sensible design, texts toward their final form. The monograph explains where that impression comes from, and shows that a large part of it arises not in the world but in the fact that we are looking at survivors.
The core idea is simple. When we study something that has lasted, we are not studying the ordinary course of events but the course of events given that it did not end. That condition changes the picture: the process acquires a visible movement away from destruction even when no movement was there. The size of this manufactured tendency can be computed exactly. It is larger the closer a system came to disappearing and the more erratically it behaved, and nearly invisible where nothing was ever at risk.
Several consequences follow, and they can be tested rather than merely asserted.
The past shifts depending on where the observer stopped. Two historians with the same sources, carrying the story to different end dates, will find different tendencies at work in the same century - and each will be right about their own data. In the calculations that discrepancy came to sixteen standard errors.
Inside a surviving system, credit for its success is misassigned in a predictable direction. Conspicuous but minor parts get more of it than indispensable ones: what cannot be done without does not vary, and what does not vary is not noticed.
And the strongest result. Natural selection turns out to be indistinguishable from sheer durability. A population under selection and a non-reproducing system with the right fragility produce exactly the same picture - not similar, but identical. The difference between the living and the non-living is real, but it lies in the count of failed attempts, which are usually gone. Hence an unexpected conclusion: going somewhere is cheap, and any filter can counterfeit it. What is expensive is staying in place against the noise, and that is what actually requires life.
The practical upshot is not skepticism but triage: the book says where the correction matters, where it is small, and gives five ways to measure it.
