As Sigit’s cadre of systematic historians travel through branching time, they practice the art of curating library furniture. Thankfully, there is usually free surplus to acquire. To a systematic historian, every study carrel and reshelving cart, covered in residual fiberdust, announces its heritage in a loud, regulated voice: Was it made before or after the Geneva Convention? Did colonialism or socialism come first? Rocket launch or space elevator?

For the purpose of recovering phylogeny, the most useful features are the least consequential and least correlated: “junk events” such as the conclusion of wars, the exchange of promises, and the migration of bodies. That’s the real reason why systematic historians love diasporas. Like a junk gene, a junk event is copied and mutated by chance without much affecting the branching rate of a history, but like a borrowed word or a healed wound, it indelibly dents the design space of artifacts that a civilization will consider and consume in its future. Mostly, people prefer their artifacts to remind them of the past either head-on or not at all. To be sure, a junk event may well hurt or please many people, and in that sense they are consequential, but what we’re talking about is the fitness of a possibility, not of a person. The ideal junk event is one that “commutes” with all other events—one that does not affect their availability.

In neighborhoods of history where the study of systematics flourishes, the value of junk events is more widely recognized, which makes them unusable. Thus, systematic historians validate an ironic sort of uncertainty principle: they know the least about their own home, the surrounding possibilities closest to them. That’s what makes Sigit’s group so precious, its mere existence.

To ease their own jobs, many systematic historians like to introduce events that, within their home time-branches, promote the preservation and proliferation of junk events. In our own time-branch, for example, the best known “junk promoter” is the advent of fashion. Of course, a junk promoter is most effective when it is introduced at an early (perhaps pre-historical) time, its home time-branch being accordingly larger. But are junk promoters themselves junk events? This is a question of research ethics that is best not raised in polite company.