Fable
The archive drew the relationships between names
The city archive used to store materials by volume. One book for caravans, one for the docks, one for taxes, one for bridges. Each book could be searched, but many questions spanned across books.
One day, the mayor asked: 'Why did last year's East Bridge maintenance affect South Port tax revenue?' The copyist found the East Bridge maintenance records and also found the South Port tax ledgers, but could not explain how the two were connected.
Someone suggested memorizing all the files. The curator shook his head. Another person suggested putting more tags on each book: East Bridge, South Port, taxes, repairs, caravans. Once there were more tags, relevant books could be found, but the relationships were still unclear.
Later, an old clerk changed his approach. He no longer only registered scrolls, but broke the city's things into individual names: East Bridge, South Port, March caravan, temporary detour, bridge repair order, tax decline. Then he drew lines between the names: East Bridge repair caused the caravan to detour, the caravan detour reduced South Port stops, and fewer stops affected tax revenue.
At first everyone found this troublesome. The old secretary said: the case file tells you that something exists; the relationships tell you why things are connected.
A few days later, the mayor asked again: 'If the East Bridge is repaired again next year, which docks and merchants are most likely to be affected?' This time the scribe did not flip through a pile of books, but walked along the lines on the map, from the East Bridge to the detour route, then to caravans, docks, merchants and tax categories.
They also discovered that certain lines on the map needed source annotations. Some came from regulations, some from contracts, and some were merely a historical event. Lines differ in credibility, and inferences differ in weight.
Later, the archive no longer only asked 'which book mentions this word,' but also 'which entities is this entity related to, what are the relationships, where is the evidence, and how far can it be inferred.' Only then did the townspeople understand: some knowledge is not a passage of text but a network of entities and relationships. To answer complex questions, one must be able to walk along this network. Later the archive hung the graph in the front hall, and merchants could also add relationships. The old clerk did not accept everything wholesale; every new line had to note its source, time, and level of credibility. Because one wrong line could skew many inferences, more dangerous than a misplaced book. The larger the graph, the more the archive valued verification and updates; as long as relationships stay alive, answers will not remain stuck in the old city's appearance.