In 1945, Vannevar Bush wrote his famous paper "As we may think" . He speculated on possible uses of current and foreseeable technology in information management on a vast scale, at a time when practically no mechanical assistance for such work existed. Bush makes no reference to any existing machine resembling a computer, but he predicts something very like one, and proposes applications far beyond those occupying the minds of people developing the infant computers at that time.
How would someone of that time design a multimedia system? It is reasonable to expect thatjust as Bush extrapolated the abilities of the technology of his daya multimedia system would be imagined as an extension of some existing communication medium. There are several possibilities: cinema, telephones, radio, and television were all well established in the late 1930s, and these are based on the same channels of sound and vision that come to mind when we think of multimedia today.
But then? The real information channel was still printfor sheer versatility, it's hard to beat a book. Indeed, text (not quite print, but close) was both the user interface and storage medium of Bush's information machine. Nevertheless, I don't think it would have occurred to me at first sight that print would be a promising basis for multimedia.
I therefore assume that it was second sight that led Christopher Isherwood to write a book , published in 1938, in which something very like multimedia is described in some detail. Two characters weave fantasies around a bookMortmerewhich they dream of, but never write.:
- As long as Mortmere remained unwritten, its alternative possibilities were infinite; we could continue every evening to improvise fresh situations, different climaxes. We preferred to stick to the Hynd and Starn stories, and to make utterly fantastic plans for the edition-de-luxe: it was to be illustrated, we said, with real oil paintings, brasses, carvings in ivory or wood; fireworks would explode to emphasize important points in the narrative; a tiny gramophone sewn into the cover would accompany the descriptive passages with emotional airs; all the dialogue would be actually spoken; the different pages would smell appropriately, according to their subject-matter, of grave-clothes, manure, delicious food, burning hair, chloroform, or expensive scent. All copies would be distributed free. Our friends would find attached to the last page, a pocket containing banknotes and jewels; our enemies, on reaching the end of the book, would be shot dead by a revolver concealed in the binding.
I do not put this passage forward as a striking example of prophecy, but I think it is interesting as showing how an author of 60 years ago, quite unconstrained by practical considerations, imagined his work being augmented by accompanying effects in different media. (It is fair to add that he had help. In a preface, he instructs the reader to read the book "as a novel," but also makes it clear that it is more or less loosely based on his own experiences and that the characters are "caricatures." He also makes clear that Mortmere is presented as a joint effort of the author and another person, Edward Upward, who, after Isherwood's death, did publish the Mortmere stories . The copy I inspected was clearly not the edition-de-luxe.)
Clearly, Isherwood himself recognized the importance of the lack of constraint ("As long as Mortmere remained unwritten ..."); equally clearly, he did not pretend to be engaged in a serious design exercise ("... utterly fantastic plans ..."). Nevertheless, here is a highly regarded literary mind toying with the idea that a work of literature can be embellished by means not normally regarded as literary, and it is interesting to list his suggestions and to compare them with current practice in multimedia systems.
- ... illustrated ... with real oil paintings, brasses, carvings in ivory or wood;
Except in special circumstances such as Star Trek , we have not yet solved the problem of matter transmission, so real oil paintings and the like are not yet available. However, we can reproduce two-dimensional pictures, and views of three-dimensional objects that we can inspect from any angle, with arbitrarily good precision. Employing virtual-reality methods allows us to manipulate the "objects," and tactile feedback, though still short of perfection, promises a realistic impression of touch in due course.
SCORE: Three topicsmatter transmission, visual effect, and touch: about one out of three.
- ... fireworks would explode to emphasize important points in the narrative;
The intention is presumably that the explosion would occur at the instant at which the reader comes to the important point to be emphasized, so again there are two topics, synchronization and dramatic effect. The effect is within reach; quite effective visual and sonic effects can be achieved with current technology. Unless the smell of fireworks is also desired (see later), that should be adequate. Synchronization is more difficult because, although quite accurate detection of eye gaze is possible, the equipment is not conventionally available in multimedia systems.
SCORE: One out of two, but with prospect of improvement.
- ... a tiny gramophone sewn into the cover would accompany the descriptive passages with emotional airs;
Once again we must consider synchronization and effect, but this case is different; the effect must persist over the period of time during which the reader is engaged with the descriptive passage. This is much easier, because the passage can be made to correspond with a single screen display, or several screens for a sustained mood, so that the period can readily be identified. The sound is commonplace.
SCORE: Two out of two.
- ... all the dialogue would be actually spoken;
Two topics again: precise synchronizationdifficult, as already describedand speechcomparatively easy.
SCORE: One out of one. (Synchronization is already covered.)
- ... the different pages would smell appropriately, according to their subject-matter, of grave-clothes, manure, delicious food, burning hair, chloroform, or expensive scent.
This time, we must have page-level synchronization, which we have covered, and smell. While we are getting better at identifying smells , a general smell generator has not so far been invented. In this case, though, we know beforehand what smells are required, so the required chemicals could be provided, preferably in highly concentrated form, and exposed at the appropriate times; similar effects have been managed with some cinema presentations. Perhaps rechargeable cartridges could be provided for friends who might read the book frequently.
SCORE: One out of one. (Alternatively, one could regard the smell problem as part of the matter transmission, in which case it is already covered.)
- ... All copies would be distributed free.
While there is still no free lunch, the Internet is a close approximation to free dissemination of information.
SCORE: One out of one. (Perhaps two out of one? Isherwood could hardly have imagined that the copies would be produced free as well!)
- ... Our friends would find attached to the last page, ...; our enemies, on reaching the end of the book...,
Apart from rather silly suggestions ("THE END. Click here if you like us or here if you don't."), we have no widely known way to determine the psychological state of the reader, but we're improving. (And fast; early drafts of this article contained the text "... it seems safe to assert that inferring the attitude of the reader towards the authors would be impossible.") In a fairly recent article , a means of determining someone's state of satisfaction by using a neural network to analyze electroencephalic signals is described, and the performance is sufficiently good for the results to be deemed practically useful, and another article  relates how a person's emotional state, detected by interpreting body-language, can be used to determine the most effective mode of dialogue.
SCORE: Half out of one, and rising?
- .... a pocket containing banknotes and jewels;
We have the matter transmission problem again. (An alternative suggestion would be to present a form authorizing a credit transfer, but that falls so far short in effect of the original suggestion that it can hardly be regarded as an equivalent.)
SCORE: Zero out of one.
- ... would be shot dead by a revolver concealed in the binding.
A matter transmitter and a self-aiming death ray? Or perhaps, to move with the times, a procedure that overwrites all accessible storage media might be considered an adequate substitute? (Methods based on the principles of the sadly familiar parcel bombs might be functional, but their brutal and indiscriminate clumsiness wouldlike the credit transferjar with the elegance of the Mortmere schemes.)
SCORE: Zero out of one.
Total score: seven and a bit out of 13. If anything, this is an underestimate, for I have written from my own knowledge and have not tried to seek out other, or more modern, information.
Considering how well, given a little license, more than half of the "utterly fantastic plans" match the facilities we have developed in multimedia techniques, it would be fascinating to know what Isherwood might have devised had he taken his design exercise a little more seriously. Would he have done better, or would he have felt constrained by the engineering of his day and produced a few tame suggestions?
Why did Isherwood succeed so well as a prophet? A plausible reason is that he simply listed what he wanted, without concern for problems with implementation. Our common humanity makes it pretty likely that what Isherwood wanted would probably be not too different from what many other people wanted, and our endeavors since then to produce the things we want have resulted in the attainment of some fraction of his wishes. (Whether getting what you want is desirable is another question entirely. I don't think I'd want to read Mortmere, even if I were a friend. It might be a tour de force, but it isn't a book, and I confessor, rather, proclaima fondness for old-fashioned, passive, off-line books that is not met by any modern, hyperactive, intrusive, distracting "alternative" of which I am aware.)
A final question: How could we do better? If we were to try to predict how a story might be presented in whatever most closely corresponds to a book in 2060, how could we do it? Can we use current technical knowledge constructivelyor would we do better, like Isherwood, to stick to "utterly fantastic plans"?
G. Alan Creak
Computer Science Department
Private Bag 92019
Auckland, New Zealand
©2002 ACM 1072-5220/02/0300 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2002 ACM, Inc.