(Note: Professor Veltman's presentation was mainly visual; an on- screen, computer demonstration of his system for universal media searching. For the benefit of those who did not see it, his remarks have been supplemented, with his consent, incorporating material from a more recent speech.)
I started writing about this in 1979 and people said it was impossible. Then, in 1986, I started putting some of this into the computer while I was Canada's first Getty Scholar, but it's really young computer programmers like Jonathon Shekter and another half dozen student volunteers who are making this vision possible.
When I was in school we were told that there was a revolution that took place around 1454. It's the revolution McLuhan talked about in the Gutenberg Galaxy: that a man named Johannes Gutenberg published the first book, a Bible.
What they didn't tell us was that it happened 746 years after the beginning of printing; that printing had begun in Korea and went on to be developed in the 12th century by the Chinese, who used it mainly to control and limit access to information.
What 1454 was about, then, was not the invention of printing but the idea that you could use this technique for giving access to knowledge to the people and not just to the civil servants to control the provinces.
Today, another revolution is taking place one which I think is greater in its import and its implications than that of printing. This revolution is linked with words like the electronic highway; that we use fibre optics. And it's already underway.
We have a link now between the Rogers Communications Centre at Ryerson University and IBM on Steeles Avenue that is pumping a gigabite per second. The people at Imax are talking about using 80 gigabites per second, so that we could have Imax pumped into the schools.
One of my teachers, a Canadian who is now head of the Vatican Library, Leonard Boyle, last year helped produce three video discs containing 162,000 pages of manuscripts. He's in discussion with the leading people at IBM to actually scan in 150,000 manuscripts of the Vatican Library.
There's a project in Washington that is proposing to scan in 10 million books. And there's a survey that's just been done in Europe of the 75,000 leading libraries of Europe that has come up with a figure of 2.2 billion (that's European billion a million million) books and a plan to scan those in systematically.
The Canadian Heritage Information Network in Ottawa, with whom we're working, has a data base of 26 million museum pieces. And they're talking about getting images of all of those available.
But the revolution isn't only about pipelining, it's about getting access to the information.
And that's why I'm doing this. I've become aware that most of our knowledge is actually not visible, or accessible, to us.
Ninety six per cent of the holdings of the Royal Ontario Museum are in the basement. The Museum of Civilization in Ottawa also has 99 per cent of its holdings in the basement.
If you go to the basements of the British Library or the British Museum you will find large crates, four meters high, with little tabs on them saying, Zambezi, 1888, unopened. You'll find hundreds of the crates, all kinds of cultural heritage which for the last century hasn't even been looked at still in crates.
The new electronic media have the enormous advantage that could bring these all together for educational purposes.
How can we make all of this accessible and how can we get it out into the classrooms? The problem is one of access.
Right now there's a war going.
On one hand we have businessmen who believe that this is a new way to make a buck who are trying to buy up the images, both of paintings and of books, on the assumption that they can make this a pay as you go thing, so that the person who can pay the most gets to see the most pictures.
Then, there is another school that says this is part of our national heritage and we have to make it accessible to everyone.
I got involved in this because I was interested in perspective. I ended up making a standard list, which amounted to 8,000 titles of text and 7,000 secondary digitry, all on file cards. I started writing about it and it got put onto the computer at the Getty in 1986. I've been working on it ever since then.
The model I'm working on is a system for universal media searching, or SUMS. We're suggesting this might be a model for what we would do if we had this great electronic highway in place.
How would we enter a data base with millions of books and images? Quite simply, SUMS would allow a person to access information by computer using a variety of different menu options at a series of levels of knowledge depending on what they wished to know. The basic system involves a series of concepts: questions, domains and levels of knowledge; levels of access, media, space, time, goals, tools and meters.
SUMS begins with six basic questions: who, what, where, when, how and why. Each of these leads to a list of persons (who), subjects (what), places (where), dates and events (when), instruction (how), and reasons (why).
Subsets of these lists are reached using the function of "limit by," which is effectively a refinement of the questions approach in order to make more manageable lists that would otherwise be overwhelming in their length.
For instance, a list of all persons and institutions would include many names that are of no interest to a given search. Hence, one would limit the lists by using the basic questions in defining a profession (who), a topic (what), a place (where) and a time (when).
If one wanted information about artists dealing with perspective in Europe from 1500 to 1800, the list would be reduced to about 1,400 names. This would include practitioners and theorists. To further limit the list, one could choose either practitioners or theorists.
Limit by has a second function. It allows one to select different views of the same material. Titles of books offer an excellent case in point. One might want to view titles only as an alphabetical list. One could type in the word, "how", and find all the books beginning with this word. Alternatively, one could ask for the same list arranged chronologically. One could further limit these by country or city, by language or key word.
Of course, if we have access to millions of documents the problem is not finding material, but rather finding too much. To compensate for this, we would distinguish between three basic domains of knowledge: pointers, objects and interpretations.
Pointers are the tools we use to get at knowledge, the materials traditionally associated with the reference sections of libraries. Objects are facsimiles of the works themselves, whether they are books, paintings, or other materials. Interpretations or analyses are the secondary literature, such as commentaries and studies which use the original object as their point of departure.
Levels of knowledge constitute a refinement in this basic strategy of identifying different domains of knowledge. For instance, pointers are subdivided into five levels: classifications, definitions (dictionaries), explanations (as in encyclopedias), titles (as in bibliographies), and partial contents (as in abstracts and indexes).
Objects form a sixth level as full contents.
Interpretations are divided into four further levels: internal analysis, which studies the subject in its own right; external analysis, which relates the book or painting to other books or paintings; restorations, where the object of study has built into it the interpretation of the restorer, and reconstructions, where this built in interpretation is even greater.
This process would entail computer aided design, animations and the potentials of virtual reality. In SUMS a simple list allows one of 10 choices, including diagram, film, photography, text, and video. And all of these media are only the push of a button away.
In order to simplify the problems of working in large data bases, SUMS reduces all of the above functions to a series of eight initial meters arranged alphabetically, namely: access, goals, levels, media, questions, space, time and tools activated by a mouse. This provides choices to meet the user's training and interest.
Because the amount of detail and information a person needs may vary with their age and education, a series of 10 different levels of access is planned, ranging from pre-school to research. For instance, in the case of geometry, a preschool child will offer definitions of basic terms, such as point, line and circle, where a researcher will wish definitions of concepts such as descriptive, analytic, projective and algebraic geometry. These levels of access would require further subdivision, as in medicine, where the needs of a general practitioner differ from those of a cardiologist, an ear, nose and throat specialist, and so on.
What's becoming possible through the computer is that not only can we have the images on the screen, but we can in fact start to see history come alive in a way that we'd never thought possible before. For instance, an architect and engineer have taken data about the town of Florence in the period 1480 to 1492 to reconstruct 10,000 houses to scale now you can virtually walk through the town of Florence as it would have looked at the time.
It would not be difficult to hook up all the great libraries, museums and art galleries the greater challenge lies in our ensuring that the technology will be available to everyone; that the electronic highway opens new doors of understanding and becomes a tool for lifelong learning.
It's my hope to see an institute that talks about the methodologies of these new technologies and gives young people the opportunity to work on a project like this.