The evolution of "Crime"

In the 1790s, Patrick Colquhoun counted 115,000 "criminals" in London by lumping thieves with scavengers, bear-baiters, and women who cohabited without marriage — manufacturing a class out of proximity to poverty. Hughes traces how this fantasy hardened: Georgian fear of the mob became Victorian belief in a "criminal class," a guild with its own argot and apprenticeships, crowned by Moriarty. The concept required no self-identification by its members; it was imposed from above, a projection of propertied anxiety onto the illegible poor. Two centuries later, the same structure reappears in PredPol and broken-windows policing — O'Neil documents how nuisance-crime data creates feedback loops that statistically conjure a criminal geography indistinguishable from a criminal race, the algorithm replacing Colquhoun's guesswork with a veneer of mathematical authority. Zebrowski's fiction completes the inversion: by the time criminal industry fully professionalizes, it merges with the legitimate order, and the justice system contents itself with "restraining the losers" — the violent and disturbed who failed even within crime — while the real operators retire on investments. The through-line is not that crime evolves but that the *category* "criminal" is perpetually reconstructed to serve whoever holds the counting apparatus.