• Contact humanITy
• ATcare (external)

• Home
• What We Do
• Who We Are
• Engage Our Services
• Contact
• Articles & Papers

Provisional but Forever: Two Faces of Internet Publishing

Return to Politics and Culture

Presentation to The Future of Electronic Publishing: The Era of Unlimited Potential

Date: 30/01/2001
Venue: Hilton Amsterdam

---

Article

0. Introduction

On the face of it there is no more final publishing phenomenon than the memorial which is, to use a lapidary phrase, set in stone:

HERE LIES JOHN SMITH
1910-2001
BELOVED HUSBAND OF MARY
AND
FATHER OF WILLIAM AND JANE

Being a lover of such local histories I go to a slightly less inflexible kind of publishing, the Riverford Parish Register and find, to my surprise, that although the date of John's death is confirmed, the date of his birth is not there; he is probably in another Register of Births but there is no method of tracing it, no personal audit trail. Here is the record of William's birth but Jane is missing.

I then consult the records of the local newspaper and find a picture of Jane, described as the "newly adopted daughter" of John and Mary; she is the lucky refugee of a civil war in the Balkans. Finally, I catch up with William who says that, far from being "Beloved" of Mary, John was a drunken tyrant on the verge of being divorced.

There, then, is a graphic example of the weaknesses of the final statement.

Not that all final statements have the same visibility. John Smith's head stone is relatively new and happens to back on to the prevailing South Westerly wind, whereas its facing neighbours are both older and more worn. Some stones have no visible writing and some of these blankly honour people who were entered into a volume of the Parish Register which was lost in a minor fire. So there are some former inhabitants of Riverford who have disappeared altogether from its graveyard and Parish records. They might be found in any number of incidental archives, a list of debtors, consumers of oysters, subscribers to a garden for the elderly, but the chances are that many of them, in spite of the initial illusion of permanence lent by a graveyard headstone, will have disappeared from our written records forever.

Let me now come a step nearer to what we would recognise as publishing. We are all familiar with the dramatic unities propounded by Aristotle but, in fact, we do not possess the document in which he set them out; we only have the word of commentators from a later age. The original record, if indeed there was one, was almost certainly lost in the destruction of the great library at Alexandria which was not, incidentally (and here is another layer of complexity) destroyed in a single, deliberative fire but was in part dismantled piecemeal by Christian fundamentalists in the decades leading up the fire. That destruction cost Western culture thousands of vital source documents, perhaps a greater number than survived from the Graeco-Roman world.

Still a step nearer to everyday publishing reality are the great stacks of the Copyright libraries. Here there are millions of books which are, in theory at least, accessible to everyone but they are only known to a variety of specialists; for most of us they might never have been. They are not in the municipal library catalogue nor on the book store CD-ROM. Some of them are mentioned in specialist bibliographies. As publishing proliferates, contemporary scholarship eclipses the historical; just as contemporary scientific and medical research consigns most material over a decade old to the historian of science or medicine, so contemporary history and criticism is much more likely, naturally, to refer to books currently in print than to those of which only one or two copies exists.

Finally, on this point, obscurity is not simply a property of the obscure. Individual manuscripts of authors and composers working since the invention of printing have been temporarily or permanently lost and some now major figures have gone through periods of almost complete obscurity. As we are re-discovering with the unearthing of Soviet bloc artistic, literary and musical treasures, war and conquest are as powerful as fashion in burying and resurrecting. The idea that Mendelssohn actually "Rediscovered" Bach is only a slight exaggeration; and if relative obscurity was Bach's fate for almost a century, what of more obscure creators?

These, then, are some factors in any consideration of hard copy publishing; items, ostensibly permanent and finished, can simply be lost. In any case, the idea that anything is ever finished is a result of the publishing medium. Most creators regard their products as provisional but the pressure of the next idea or the next grocery bill or the next contractual deadline force them to abandon what they would prefer to revise. In fact, as media have become more plastic and the production of revised editions easier and cheaper, more creators have devoted an increasing part of their creative life to revision, re-working and commentary on their own work.

Before concluding my introductory reflections I want to add one more piece of evidence. The Prefaces of many books conclude with a deceptively simple location and date but many of the great books on which our culture has been built were written over long periods. Take, for example, St. Augustine's City of God which was written partly in reaction to the sack of Rome in 410 AD. It was written against a background of political turbulence not seen in the Empire since the Imperial wars c68 AD; those Sections written at the beginning did not share a political context with those written at the end. Nor was St. Augustine's theology so settled that the book can be taken as a coherent whole; as he wrote and read he changed, but dictation to scribes and the final production of a manuscript did not lend itself to intratextual criticism, interpolation and cross reference. Revisions are made and frequently acknowledged and where this happens are we, as we would in considering a Last Will and Testament, to take the chronologically last statement on a subject as the authority over-riding all others? What about the cranky codicil? This is, of course, not simply a problem with ancient manuscripts. One of Augustine's intellectual descendants, the Rabbinical Karl Marx, left behind a host of disputes about the sequence, balance and context of his pronouncements. In other words, the idea of a book being realised in an instant of time which makes it internally coherent, like the snapshot from a camera, is an illusion but one which has coloured much of our intellectual life. Only perhaps with Vico did we discover the idea of historiography and only perhaps with Michelet did we discover the idea of chronology as an accelerating force in narrative but we are naive about medium. If the phrase with which I started, "Set in stone" is one measure of supposed solidity then "IN black and white" does not lag far behind.

With these ideas as a background, I want to look at the future possibilities of Internet publishing, considering the following five aspects:

1. The Nature of Hypertext
2. The Permanence and Provisionality of Publishing
3. The Total Market Concept
4. Customisation - Authors and Consumers
5. Migration Tools and Packaging.

These roughly correspond with but are wider than the initial topics on the programme.

1. The Nature of Hypertext

One of the really strange intellectual puzzles of the new information age is that hardly anybody using HTML actually knows what hypertext is. I agree with Umberto Eco that Hypertext, as an idea, was first formulated by the philosopher Leibniz just over 400 years ago. The essence of the idea is that a piece of data can be classified in more than one way so that, for instance, an event such as the Nativity in the Christian New Testament may be classified under History, Theology (both Christian and Islamic), anthropology and mythology. The problem which Leibniz faced was that up until that time the 'tree' structures of classification, attributed to Aristotle, only allowed phenomena to be classified in one place; so an event or phenomenon could only be located at one point, and that point might be an extreme twig at the end of a bifurcating tree structure. Strange, then, that the most common information structure in the world of hypertext is still the tree. The only saving grace of the digital tree is that it takes advantage of the first great fruit of Leibniz' work, the worthy cross reference.

There are three phenomena of hypertext which are particularly relevant to Internet publishing:

• Internal Navigability
• Reciprocal Linking
• Brokering.

• First, internal navigability is a crucial phenomenon in information access but it is not well served by tree structures. The structure I currently most favour is the cylinder bent round until its two end cross sections are joined in what looks like a doughnut without a hole, at whose centre point there is one node which can be reached from any point of the structure with one step. This offers an infinite series of circular plates, the cross sections of the initial cylinder. Each such plate can be divided into any number of segments or fields or each can simply accommodate an infinite number of nodes on its circumference. Common nodes on the circumference of each plate can be connected by lines running along the outside surface of the structure. As relationships between subject matter varies, each plate can be rotated on its own axis to create different lateral relationships. In such a model the objective of taxonomy is not to isolate but to join. The golden rule for publishers is that if you are offering a pay per consume service, you make no money out of search time and the longer the search the more likely it is that the potential customer will give up without buying anything.
• Secondly, the key to success is not just links from a piece of hypertext but links to it. No matter how good your dispersal mechanisms they are redundant if you have no arrivals to disperse. This is a good instance of working out where to compete and where to collaborate. The idea of discrete publishing is anachronistic. Your primary source of income will arise from content quality and navigability, not from the erection of barriers. We will return to this later under the heading of Copyright.
• Thirdly, the essence of Internet publishing is brokering as well as building. If you can get paid for delivering passengers to a piece of information on an alternative site there is no point in creating and holding that piece of information yourself. The amount of duplication in cyberspace because this principle is not understood is simply immense. We still largely think of the Internet as a method of providing a channel for client/service traffic where each end of the channel is sealed.

This is not to say that the only mission for the Internet is effectively to connect, collate and re-segment all existing data into different kinds of information. Of course we will need to create masses of new content but we can only do this effectively if we understand how to marshal what we have already got, leaving us free to be creative instead of painfully re-working for ourselves what has already been achieved by others.

Let me give you a simple example. One of the most important requirements for text only websites is that graphics should be described. This is not only of importance to blind and visually impaired people accessing the Internet from a PC, it is also important for people with legacy PCs and for those wishing to access the Internet via a WAP phone. Great paintings are notoriously difficult to describe so the obvious route is to link graphics titles with art gallery catalogue archives rather than trying yet once more to describe the smile on the face of the Mona Lisa. You save time, provide a better service and, on that basis, can be paid for the routing rather than the direct provision.

2. The Permanence and Provisionality of Publishing

Next I want to talk about the paradox of permanence and provisionality. Books and magazines look permanent both in their form and content but, as we have seen, they can disappear completely and the idea that their content is somehow final, beyond amendment, is an illusion. The products of Internet publishing are, on the other hand, potentially eternal and absolutely provisional. These unique properties are at once a blessing and a curse, a blessing in that good work is assured of an unlimited future, a curse in that, no matter how competent the initial exposition, such work usually loses relevance over time. There is nothing so abject as an out of date website.

Permanence and provisionality are, in some senses two sides of the same coin so that, for instance, permanence and provisionality both require updating, so here is a basic list:

• Templates
• Weighting and Routing
• Chronology and Typology
• Updating
• Databases
• Migration and Transformation
• Tools.

Templates

You will save yourself a great deal of trouble if you design an inclusive rather than an exclusive template. At the very simplest level, for instance, this means that series should be capable of internal expansion. Using eight factors in numerical order is no use if you want to insert a ninth between the fourth and fifth at a later date.

Weighting and Routing

One of the most unhelpful aspects of post modernism is its reluctance to assign value. Now I may not wish, on that basis, to assign value to content in the sense that I think what I have to say is more important than what you have to say but if I am publishing on my own site then I am entitled to make a distinction between what I say and what other people say about what I say. It ought, therefore, to be possible for a user to follow a route through what I say, distinguishing that from criticism and metacriticism, no matter how valuable.

Chronology and Typology

It is vital to assign a chronological date to all data and to assign typology or, in other words, to assign some status. Traditionally this information was carried on the page backing onto the title page of books, showing the details of various editions of a text. Nowadays this is often not practised by website designers so information which is intended to be temporary and updated is presented as if it were timeless and had the status of an unchanging phenomenon. In order that what you say is treated as timeless it must be allocated a time.

Updating

If you cannot regularly update information yourself, use data mining software with clearly defined chronological criteria for producing possible information to enhance an initial presentation. To take this presentation as an example, I will post it to my website but as long as it stays there I will make sure that relevant future research commenting on my ideas is cited; in this way I will chart whether what I have said has turned out to be helpful or largely misleading. If it turns out that the latter is the case I will reach a point where this document is no longer of any real use to publishing analysts at which point it will be consigned to an archive. This means that I have drawn a distinction between directly relevant information which is still subject to succeeding layers of criticism and archive material which is only of real use to historians. I can best illustrate the distinction by referring to medical textbooks which go through gradualist revisions, reflecting current and recent controversies but there comes a point when so many new discoveries have been made that such books are only of interest to the historians of medicine, as opposed to practitioners; we might be fascinated by the role of the leech in medicine as an historical phenomenon but we prefer modern drugs. We ought to be meticulous in making this basic distinction.

Databases

Although the distinction is narrowing between word processing and databasing, information should be made available in a database format to facilitate searching and modular updating. Again, remember the taxonomical point about the same data being entered under different classifications.

Migration and Transformation

As technology changes, tools will be required to convert formats to new platforms and new environments. New systems may be backward compatible but they may not or they may only be so for a limited period; thus, a new operating system may provide for legacy migration to one or two of its brand predecessors but not to three or four.

Tools

There is no longer a distinction between data and the tools which realise its access. As we will go on to see, it is not the initial authorial intention that counts but the consumer's realisation of it and that may require the use by different individuals of different tools to consume an end result not identical in every case but as near a match as possible between the author's intention and the consumer's capacity.

Of course, you may not wish to go through all these procedures because what you are actually doing is setting off a publishing firework that flashes into glorious multicolour and then fades. You might be promoting a one-off event, a simple phenomenon. There are those who believe that the whole corporate world is going to be a series of short term projects where brands appear and disappear within a couple of years; that our industrial enterprises are going to look like provisional task forces and ad hoc project teams. That is the kind of vision which stirred Internet pioneers who thought that the civilian Internet would be a haven for anarchists where there would be a glorious free-for-all. Well, my heart may be with Howard Rheingold but my head is with Commissioner Bangermann. The .com boom was based upon naive economic models. We will need some more robust structures but they will all follow a simple rule that if you can't get a return on investment the investment will dry up.

So it is with publishing and brands. There is so much ingenuity and effort required to establish a good brand it will take some time for investors to receive a reasonable return. The theory of provisional branding is just that, a theory. We are currently in a phase of febrile start-ups and buy-outs but remember that in 1900 there were 300 automobile manufacturers in the United States and now there are three. The same evolution will happen in cyberspace but it will be much more rapid. We have already seen that anyone can become an Internet Service Provider or establish a broadcasting system but without content to communicate, communications systems are not of much use.

There may, of course, be instant brands and associated information such as that to promote a one-off event but we must be careful not to confuse the temporary with the permanent. Long-term publishing deserves a massive amount of initial care and planning simply because its essence is that it is long term and, by and large, incremental; this means that its requirements are very different from temporary publishing. What is more, the permanent publishing strategy requires a careful market analysis so that it can achieve the maximum return on investment and it is to that which I now turn.

3. The Total Market Concept

The first thing to grasp about Internet publishing is that it is rapidly converging with broadcasting, not just because publishers are reaching wider, general markets but also because broadcasters are continually narrowing their offerings. As a term, "Convergence" slips off the tongue easily enough but what does it really mean in the case of selling digital information?

Leaving aside the actual limitations of media such as magazines and terrestrial television, let us start with the essentials of the offer from these traditional media.

Publishing has almost entirely relied upon directly serving one or more, clearly identified, market segments. Customer groups obviously overlap so that individual readers have a variety of identified interests; this is what makes advertising possible. There is only a very limited number of sales for football boots but a much wider market for lager. Still, the essence of the offer is that you know your market, its intellectual and emotional level, the degree to which it can handle complexity.

Terrestrial broadcasting, on the other hand, particularly in Europe, has assumed that its main function is to make an offer of content which is optimally accessible. I remember with a deep sense of pain those childhood Saturday evening entertainment shows with a slightly risqué comedian for dad, dancers for mum, a ballad singer with an accordion for granny and a pop group for me; we all hated each other's pieces but there was no alternative. Too much ghetto broadcasting would drive away the mass audience.

It is a mistake to see this purely in terms of medium. An all sport television channel looks more like publishing than broadcasting and many major newspapers look like broadcasts.

What we are now entering upon is an era when we can look at our content precisely in terms of the breadth of the market segment we want to satisfy. Our basic content can be part of a broadcasting offer but it can be progressively narrowed, through middlecasting, to the narrowest of narrowcasting, to an audience of one.

So if you are shifting from a fairly narrow publishing operation to a much broader Internet environment, what do you know about the general audience that might come across you, quite deliberately, as the result of a search or by a complete accident? What do you know about the navigation skills of the population; does it matter that maybe a quarter of your audience will not know left from right? How big is the lexicographic range of various segments of your audience? Is it true, for instance, that many people have very small general vocabularies but comparatively large specialist vocabularies? Think about people in a bar who hardly have general words other than expletives but who have a soccer vocabulary, excluding proper names, running into hundreds of words.

Of course most information publishers/broadcasters are not bound to consider the whole market as an audience but this is a good place remind you of one point. Governments have to regard the whole population as an audience, whether people are filing tax returns or responding to a consultative document. In the new digital age governments will not be able to discriminate against citizens in information provision simply because they are not very clever; you don't have to be clever to vote.

Publishers, of course, are not so bound by the universal ideal but, still, if you have information it is important to exploit it to maximum advantage and, like the retail of children's toys, you never know with Internet information what is going to be a big hit outside, of course, sport and pornography. What you do know, however, is that the information you provide has a potential unlimited life and with relatively simple tools can be made accessible, as well as available, to very wide audiences. These two concepts - accessibility and availability - are often treated as if they were interchangeable. If a government puts a 500 page technical report on its website it is universally available but hardly universally accessible.

If you think again about your market, you will immediately recognise that it consists of people with a variety of physical and mental differences which make access to multimedia more or less easy. What you really need is a sophisticated demographics of IT and functional limitation but there isn't such a thing. All that you can do is to make some guesses about the four broad categories of limitation:

• Mental and Cognitive
• Physical
• Hearing
• Visual.

There are many sources of specialist advice about the last three so I won't say more than a few words about them. Here are some summary points:

• The more mild the more sufferers
• Disability spectrum
• Elderly rather than children
• Often well off

The milder the syndrome the more people suffer from it; so there are many more people who cannot read standard print than there are blind people; it therefore follows that

There is no sharp line between able and disabled people. Conditions run from severe to mild along a spectrum. Disability categories are administrative or medical but they have nothing to do with the consumption of goods and services.

In spite of charity fundraising images, most disabled people, particularly those with mild conditions, are elderly not children.

The sector of the population with the highest level of disposable income is that whose children have finished their education but who are not yet liable to high medical bills. They are often mildly functionally limited in respect of IT. The idea that society is split between a natural class of IT enabled people who are relatively well off and an IT disabled class which does not have any money is bogus. This false dichotomy is potentially the largest single factor in artificially lowering IT profitability, whether you are thinking about hardware, software or information.

But the first umbrella grouping - those with mental or cognitive functional limitations - is the largest but the most difficult to come to terms with. For a start, it is very difficult to persuade adults to undertake literacy tests when they fear they might be found less than adequate; and people are very good at disguising large cognitive failures by substituting memory and guessing for actual comprehension. There are problems, as I have said, with lexicographic range, syntax, navigation, self-expression, and so on.

The key to reaching out from a relatively narrow audience to a much broader one is customisation, and that is my next topic.

4. Customisation - Authors and Consumers

As I remarked earlier, tools are not external phenomena which are brought to bear on information, they are an integral part of it. Customisation is largely a matter of applying tools to information but it is immediately obvious that in some cases the application of tools changes the information itself. The simplest example is that if you double the size of every element on a screen the information has to be divided between a prior and a succeeding screen. It is no longer presented as an integral array but has been re-created as a simple, two step sequence.

Until now the problem has not so much been designing tools for customisation, it has been that of establishing and saving settings so that customisation is not a tedious prerequisite to every session with a PC. There was a time when it was impossible, using Microsoft, to establish, save and move settings but that can now be achieved in NT; and there will be a time when intelligent software will provide each user with a default set of characteristics based on its experience of each user; but we are still in transition from almost total inflexibility to heuristic middleware.

Some tools, such as those for volume control, colour selection and contrast, magnification and speech to text translation, are relatively sophisticated but some aspects of language engineering, for instance, are still very primitive.

What I have done, to make the issues relatively simple in presentation, is to put the four groups of functional limitation on a horizontal axis and information systems requirements on a vertical axis; and I have then filled in the boxes with practical suggestions, some of which relate to hardware (such as the ease of turning an appliance on and off) but most of the requirements can be met with tools.

humanITy’s Accessible Information Matrix (AIM) Version 2.0

Now the key objection to this approach is that any customisation, even through the use of a temporary file, somehow perverts the initial intention of the author of the intellectual property. Somehow, if she set the text in blue on a crimson background her integrity is being violated if we change it to violet on green. Well, perhaps in a very limited number of cases there is a point to that; I doubt someone as particular about the quality of light as Monet would want his pictures re-tinted. But we really have to get away from the absurd 19th Century idea of art as in some way sacred and separate from the rest of human existence; we have to think of information provision as a craft, the main point of which is the ability of the consumer to grasp as much as possible of the author's content. Let us think of Luther, not Goethe, of Telemann rather than Berlioz, of M. C. Escher rather than Van Gogh. There is ample room in our cultural space for highly individualised statements which can in no way be amended without imposing this ridiculous rubric on all forms of intellectual property. If you follow the development of Haydn String Quartets you will notice that as well as being musicological it is dynamic; the later quartets written for public performance in halls accommodating up to 800 people had to be very different from those intended for a salon soiree. Today's self-styled purists would be staggered at the extent to which Mahler altered the instrumentation of his symphonies to accommodate them to different concert halls. Outside Christian scripture, examples are not so easy to find in literature but they are, of course, abundant in the mining of literature for film.

If we are going to rely upon simplified realisations of intellectual property then it is much better to achieve this through tools or contracted intermediaries rather than through partial intermediaries without any responsibility to the author. In insisting on a sharp distinction between the creator of intellectual property and its manipulation by mass media we are losing the opportunity to think more clearly about preserving the essence of the author's intention in spite of adjustments to the means of communicating it.

At the heart of our agonising is the traditional idea of copyright. The new European Directive on Copyright is, I fear, out of date before it is even promulgated. It has simply examined the old rules of copyright for traditional media and has ignored aspects of content alteration which are made possible, and sometimes necessary, by multimedia, such as:

• Simplification, e.g. through the use of language engineering tools within a language
• Transcription, e.g. putting text into braille or automatically generating subtitles
• Elucidation, e.g. transforming simple key words into hyperlinks
• Enhancement, e.g. adding graphics to a difficult text only document
• Interpretation, e.g. providing parallel commentary
• Interpolation, e.g. creating marginalia within a file.

As an author myself, I would be prepared to allow anything to be done to my work along the above lines as long as any successor file has a route back to the original which I created and, therefore, implicitly acknowledges my work and the contribution of intermediaries.

If authors wish to operate in the public domain in a human rights context they should not be in the business of telling people how they may and may not access content put into the market for consumption. Publishers should think very carefully about concentrating on what the consumer consumes rather than what the producer produces. Once you have made a deal with the producer, after all, your job is to make as much return on your investment as you can through the consumer and at the moment that is being made more difficult by intellectual property rules which combine the inflexibility of hot metal printing with the worst excesses of high romanticism.

As for the stubborn tendency of IP creators and lawyers to insist on a new regime for every new medium, this is an obstacle to growth and exploitation and it is to this which I will finally turn.

5. Migration Tools and Packaging

If the laws of inheritance were as strong as the laws of copyright, Homer would be haggling over the Webcasting rights of the Odyssey. Having struggled to obtain the parchment rights and the moving type rights, the film and television, radio and recording rights, the trustees of his estate would now be on the lookout for rights in every new format until the end of time.

If publishers are to do justice to the technical possibilities of varying the range of audience (broadcasting, middlecasting, narrowcasting) and providing a wide array of customisation and migration tools, then it follows that intellectual property creators must either accept a lump sum advance for their work or accept a percentage of its consumption proceeds on a medium and carrier free basis, or a combination of the two. There must be no more dissection of rights and royalties every time Sony Electronics creates a new way of storing and delivering information. If publishers create the means for longevity, then authors should make that longevity as trouble free as possible.

We have already seen bloody warfare over film rights and, quite irrationally, audio rights; but, whereas Hollywood employed scriptwriters the new media will simply require migration tools. There is a trade off between building information to last and relaxing the rules on migration and re-packaging.

End of the shortest section of this presentation.

6. Conclusion

Finally, I want to draw some of my threads together into a warning and a challenge.

The warning is this. The United States media policy which has so far been implicit will, under George W. Bush, become explicit; that policy is to use the World Trade Organisation to force a global market in information. The size of the USA internal market and the increasing use of a kind of English give it an overwhelming advantage. If we are not careful, what happened to meat yesterday will happen to media tomorrow. In resisting this pernicious campaign which equates culture with corn, the French were absolutely correct. If we want the kind of multimedia I have described which is:

• Rich
• Inclusive
• Long Lasting
• High Quality.

Then we will have to make some sacrifices, greater than the cashmere industry molestation in defence of America's banana plantocracy. For once, the European Union will have to raise its sights above trade, as will we all. There is an Anglo Saxon neo Liberal economic model and a European social model and in terms of labour and land, mortgages and motor cars, we may take our choice how far we tend towards one or the other; but in culture, in honourable and creative publishing, we must stick to the European social model, with its roots in public broadcasting and its strength in diversity, in small companies and big ideas. Because, ultimately, like money, the bad drives out the good and, if that happens, we in Europe are bound to lose.

Return to Politics and Culture