I went down the rabbit hole on this 🙂 Hours well spent

In the book I found here, at the very end the author mentions early work on computers to index and manage topic-based access to sources. I was stunned that so early as 1961 this was already in books for the general public. The system was called SYNTOL (Syntopical something). I search again and again and found a single book referring to it https://www.amazon.fr/Automating-Linguistics-History-Computing-English-ebook/dp/B093KKR99S

Lucky enough I wait a day and thought that it could exist in French. It does and is open source. There you find precursing ideas of the modern use of NLP to solve this problem. It solves the heavy work of numbering and keeping pointers and collections. I use a similar philosophy in Kneaver but with modern NLP.

Bottom line in 1961 the state of the art on the topic was already a lot ahead. What was missing is Internet, personal computers, and access to digital form research. Mid-1980s that was available and cards could be shelved. Now we have AI-based NLP are it becomes reliable.