A group at Harvard is trying to do online what would have been impossible before the Internet: compile a library of nearly every book available (see “The Library of Utopia.”)

The project is ambitious, but the idea, or something similar to it, is nothing new. One examination of the concept came in a May 1961 article called “How Will You Obtain a Book in 2000 A.D.?” The piece covered a speech by John G. Kemeny, the chair of Dartmouth’s math department (later the college’s president and a developer of the BASIC programming language), who advocated what he called the “automated national library.”

This library for 2000 A.D. would start with 10 million volumes and possibly grow to 300 million within the Twenty-first Century. Each page of each “book” in it would be stored on possibly a square millimeter of tape. The library would serve 100 or more universities, each of which would have a multitude of viewing screens. Using the library would be similar to making a long distance telephone call: When the patron dialed the correct code number, the reference he wished would appear on his viewing screen.

Of course, one potential problem would be that multiple people would request the same item at once. Kemeny anticipated this: under his system, anyone asking for a particular book wouldn’t get the original volume but a copy.

The item asked for would be transferred from the storage tape to a projection unit and sent, conveniently magnified, to a tape in the reading room of the person desiring it. With $10 worth of tape, a customer could have a 10-volume collection of personally selected items; and if he tired of them, he could simply erase the tape and have a new set of books flashed to him from the master tapes in the central library. One day of each month, however, each branch of the central library might have to be closed for updating, extension, and repair of its master tapes. This, of course, would be only a slight inconvenience.

Kemeny felt that the libraries of 1961 were already “practically obsolete” and would be “useless for most purposes by 2000 A.D.” But who would pay for his elaborate upgrades? He conceded that the project might cost $1 billion, but he argued that this might compare favorably with the cost of a hundred separate universities maintaining a hundred libraries.

The library of the future will have to make use of automation. There is no conceivable way in a library of several tens of millions of volumes that human effort could locate an item in a matter of minutes … Storage methods must miniaturize books and put them on a medium easily handled by machines.

This sounds vaguely Internet-like, but Kemeny wasn’t predicting the Internet, exactly. He made no mention of anything like hypertext, for instance. But he did foresee the utility of being able to retrieve information over a phone line and view it on a screen from many miles away.

I am particularly attracted to the prospect of combining this automated library with machine-search … It is possible that, even with all this elaborate mechanization, information retrieval will become hopeless in 100 years—but without mechanization we won’t stand a ghost of a chance.

Kemeny also hit on what would turn out to be one other benefit of the Internet age: the ability to get what you need without leaving your chair.

I find the concept of such a library very attractive. I am basically a lazy person. I would like to sit in my office and have access to a book with no more trouble than calling a friend on the phone long-distance.

Timothy Maher is TR’s assistant managing editor.