Fable
The inquiry window learned to check the evidence before speaking
The city's inquiry hall collected manuals, contracts, and old notices from various shops. When customers came to ask about warranties, returns, or sizes, the clerks had to find evidence first, then answer.
In earlier years, librarians answered from memory. They were fast, but often treated old rules as new ones. Once a customer came to the shop based on an old return period, the shop refused the return, and the two sides argued all the way to the government office. Only then did the head librarian require every answer to point to a specific page. The first time the head librarian mandated "must cite pages," the librarians merely attached page numbers to the end of answers. Customers saw a source and assumed it was credible, but the cited page could still be wrong.
The first version of the new rules was also unstable. The apprentice found a page containing the keyword and copied it directly, so when asked "can rain boots be returned," they only found the ordinary boot clause and missed the rain gear appendix. The answer seemed to have a basis, but still didn't answer the question. After that rain boot dispute, the librarian reviewed and found that the keyword "boot" led them to the ordinary clause, while the customer's real point was "returns and exchanges after use in rain."
Later the inquiry hall changed to three steps: first turn the guest's question into searchable clues, then retrieve relevant pages from the shelves, and finally organize the pages into an answer that can be understood. If evidence was insufficient, they said plainly that it couldn't be found and didn't pretend to know. The three-step method has boundaries too. If the manuals on the shelves were outdated, or the relevant pages contradicted each other, the clerk had to flag the conflict first and couldn't force a smooth-sounding answer.
The curator also required answers to address the question at the start. If a guest asks whether something can be returned, answer yes or no first; evidence follows. If either finding the material or organizing the answer is missing, the guest does not receive real help.
The inquiry hall also kept an "I don't know" sign. If evidence could not be found, the clerk would rather ask the guest to wait than fabricate an answer from memory. The credibility of answering based on materials often comes from admitting that the materials are insufficient.
Later, when the inquiry hall added new manuals, it also cut them into searchable chunks first, marked with version and scope of application. If the materials are not well organized, the first step of the three-step method will go off track; the reliability of the answer depends on the bookshelf and the question being reliable together.
The reputation of the inquiry hall slowly returned. Being able to answer questions is not just about speaking well, but about finding the right materials, reading the right context, and letting guests see where the answer comes from.