Memex Research Project

Produced in collaboration with Ryan Dodd.

Precursor (What came before Memex?)

The Only technology before Memex was the Mundaneum (1). Originally called the ‘Universal Bibliographic Repertory’ it was formed by Paul Otlet and Heri La Fontaine in 1895. The aim of the Mundaneum was to have a “City of Knowledge”. The links to various documents such as magazines, books, and pictures would be written on cards and archived in small drawers inside cabinets. Someone at the institute could be contacted by telephone or telegram where a query on a particular topic could be looked up on the cards. Retroactively the mundaneum has been cited as a precursor to the web as a way to access information. (7)


Prior to inventing the Memex, Vannevar Bush had a background in electrical engineering. During World War I he worked for the National Research Council to create new forms of weaponry one of which was a “a device that would use magnetic fields to detect submarines”. During the 1930’s Bush also developed the ‘rapid selector’ a device which was able to archive documents onto microfilm. Documents that had been archived onto the device could then be projected onto a screen and switched between at a fast rate. Neither of these inventions where able to catch on as they suffered from various technical flaws. (8) Bush put forward the idea for the memex in 1945 in an article entitled “As we May Think” in the Atlantic. (2)


Method and the Theory behind Memex

Bush had come to the conclusion that information collected through years of scientific study had become inaccessible due to the sheer volume of said information. “We are no longer able to use what the science finds out.” (2) In order to combat this he proposed the memex a desk which could store documents in the form of microfilm. Bush wanted to have users create links between separate documents by associating each document with a code that could be referenced. The idea behind this was to create a trail between documents so that users could find out more information associated with any one particular document by referencing what was related to it. (2) This concept can be seen as an early forerunner to the World Wide Web, specifically to search engines such as Google whereby the user can find documents, pictures, etc based on metadata that can be referenced through a search of related keywords. 

The pros and cons of Memex

While Bush never actually made the Memex, if he were to have attempted it, a huge variety of technical challenges would have presented themselves. One such limitation would have been the use of microfilm as a storage device. While it was useful to a certain extent, it was not an infinite method of storage. If Bush investigated the alternatives, he would also have come across limitations from the proposed use of photoelectric components, and then crystalline memory. It is widely thought that due to the technical limitations if it were produced, that the function and design would have been greatly deviated away from Bush’s original intentions. (5) Despite this, and coupled with the fact it was never built, the Memex is a theoretical machine, and was ahead of its time. The Mundaneum was more indexing, whereas Bush’s Memex recognized the importance of links between data. It was the first visionary description of the potential use for information technology. (3) This was a unique vision, and even though it wasn’t built, the idea is the key point here.


What followed Memex, and whether it had any significant affects?

Memex led to the eventual use of hypertext in the World Wide Web (widely recongised as the theory of this, even if not the implementation). In an attempt to quantify this, academic Linda C Smith traced the links between hypertext implementation and theories presented in Bush’s As We May Think, finding that modern hypertext was indeed directly traceable to the theory he presented. (4). Smith’s research is the closest thing to proof that Bush’s Memex is the main precursor to hypertext. Bush also made a number of predictions in his article, one being of new forms of “ready made” encyclopedias (one could take Wikipedia as an example of this) that could simply be “dropped into” the Memex. With the staggering growth of such services on the web, his prediction, in a way, proved right. (2). The latest link back to the ideas of the Memex is a very recent one. DARPA (Defense Advanced Research Project Authority) began in 2014, a project (named Memex, assumedly in tribute to Bush) to provide a search capability for the dark web. This uses Memex theory as inspiration, and was the first major project of its time, showing how the impact of Memex can be traced up to the present day. (6) The concept of the memex specifically that of creating trails between similar documents and pictures may have inspired Ted Nelson who came up with the concept of ‘hypertext’ for use with ‘Xanadu’ as a way of sharing documents. (9)


  • The Memex was the idea of Vannevar Bush, an American electrical engineer, who outlined his version in an article for Atlantic Magazine, entitled “As We May Think”. He was concerned about the sheer amount of scientific knowledge that society was amassing.
  • Before Memex, there was the Mundaneum, that had limited functionality
  • The Memex was a desk-style piece of furniture, that stored microfilm based documents. Through a “system of levers” links could be created between document metadata, that would allow you to reference one document from another.
  • While the Memex was never built, the theory behind it has been recognized by many a precursor to the hypertext system used in the World Wide Web.
  • However, from an engineering perspective, the Memex had no blueprint or feasible method of operation, which is an obvious drawback.
  • In the latest development, DARPA have named their new attempt at a search engine for the dark web, Memex.

Interesting Memex Fact:

Memex was a name, to quote Vannevar Bush “coin[ed] at random”:

“Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and to coin one at random, “memex” will do.” (2)


One could speculate that this shows he was not massively focused on the mechanics and the small details of the machine, more the overall theory and what it represented.



  1. Carr L. Precursors to the Web. Web Science Lecture; 2015 Sep 30.
  2. Bush V. As We May Think. The Atlantic. 1945 Jul;6.
  3. Vannevar Bush and Memex [Internet]. Living Internet. [cited 2015 Oct 1]. Available from:
  4. Smith L. Memex as an Image of Potentiality Revisited. 1991;265.
  5. Barnet B. The Technical Evolution of Vannevar Bush’s Memex. DHQ: Digital Humanities Quarterly [Internet]. 2008 [cited 2015 Oct 1];2(1). Available from:
  6. Zetter K. Darpa Is Developing a Search Engine for the Dark Web. WIRED [Internet]. 2015 Feb 10 [cited 2015 Oct 1]; Available from:
  7. Eric Pfanner. ‘Google to Announce Venture With Belgian Museum’. NY Times. [Internet] 2015 March 12.
  8. Internet Pioneers. lbiblio [Internet]
  9. Gary Wolf. ‘The Curse of Xanadu’. Wired. [Internet] 1995 June.


Leave a Reply

Your email address will not be published. Required fields are marked *