Ranking as an Index

The Index worked by making a book commercially unavailable rather than by destroying it. The book existed. It was simply not printed, not stocked, and after two generations not known. This archive has already argued that this is the standard shape of suppression: not destruction, but removal from the paths people actually travel.

Search ranking performs the same function with better coverage and no published list.

State the comparison precisely, because the loose version is a complaint about censorship and the precise version is more interesting. A document that ranks below roughly the tenth result for every query that would reach it is, for practical purposes, in the same position as a book on the Index: it exists, it is not illegal, and almost nobody will encounter it.

The differences run in both directions. The Index was published, so anyone could see what was on it and seek it out deliberately, and people did. Ranking is not published, changes continuously, differs by user, and cannot be enumerated. There is no list to consult and no way to determine what one is not seeing.

But the Index was also enforced by an institution that believed it was protecting people from error, and ranking is optimised for engagement and commercial return. Neither of those is truth-seeking. It is not obvious which produces a worse distribution of attention, and that is not a rhetorical flourish; it is genuinely unresolved.

The mechanism has a second-order effect the Index did not. Because ranking is the primary route to readers, the material itself is shaped to rank. Publishers, writers and institutions produce work in the forms that perform, which means the ranking function is not merely selecting from what exists but determining what gets made. The Index had no comparable influence on what authors wrote.

None of this requires anyone at any of these companies to have suppressive intent, and this archive should be careful not to imply otherwise. The systems are optimised for a measurable objective. The distributional consequences follow from the objective, and the objective is not concealed.

That is the argument, and it is a familiar one by now. The suppression that matters is usually structural rather than intentional, and structural suppression is harder to see because there is no decision to point at and no one who could reverse it by changing their mind.

The practical response is the one the archive keeps arriving at: redundancy and deliberate routes. Feeds you subscribe to rather than feeds assembled for you. Bibliographies. Following citations backwards. Physical libraries. Asking people. Each of these is a path to material that a ranking function does not mediate.

There is a reason the earlier entries in this series are about mechanisms rather than villains. Once you can see the mechanism, the current instance is recognisable without needing anyone to be guilty.