This report is also available as an Acrobat file.

Contents
Authoring and Design for the WWW

PRINCIPLES AND THEIR IMPLICATIONS

The uses of hypertext

Dynamic behaviour

Even where text alone is used, interactive documents can contain other kinds of interactivity than straightforward hypertext. For example, an interactive table of programme modules could in principle allow the user to click on a module and automatically highlight any prerequisites and/or progression opportunities appropriate to it. However, it should be noted that this presupposes a substantial amount of development work by staff, a higher level of skills, and suitable authoring and delivery tools. Again a staff-development issue is raised. It is not suggested that many staff members could or should explore the intricacies of creating multiply interactive documents but perhaps they should be sufficiently knowledgable to specify requirements for such elements to experts.

Another crude but important form of interaction includes the filling-in of electronic forms, which allows students to actively interrogate the information, and to become contributors to that information themselves.

An unquantifiable motivational aspect may be present in interactive documents, simply by virtue of their interaction. It seems possible that rather as one gets more out of talking to responsive students than to a room full of blank faces, so interactive documents by their very responsiveness may seem more congenial than an inanimate paper document. This motivational aspect has received a lot of attention in relation to interactive academic learning materials, but is also applicable to institutional information.

Mixed media

The principle of using interlinked electronic interactive texts is easily extended to other media. This means that the most appropriate medium is available for each element of the communication task. Numerical information concerning the proportions of credits for component modules might be animated in order to show their accumulation over time. The user might be able to try what-if' combinations of modules and receive instant feedback on legal and illegal combinations, preferred routes, progression points, etc. A welcome to the Set might be given as a digital video presentation by a member of staff, adding an element of human warmth and engagement which would be more difficult to achieve in a paper document.

What do images do that text does not? It has to be said that there is little objective evidence for the superior communicative power of multiple media as against text supported by well thought-out illustrative graphics. However there is at least likely to be a motivational element in an interactive multi-media document, which will encourage the user to explore more deeply than might be the case with a paper handbook.

The user's experience of hypertext

While hypertext invites document authors to make non-linear structures, they are not compelled to do so. The Web can be used as a place to deliver a document which is traditionally linear and contains no links at all to other documents. Nor does it impose any particular structures the choice lies with the author and designer. They can organise the nodes and links of hypertext documents in different ways in order to suggest to the user a variety of different structures. Examples include:

a linear document with hypertext links to entries in a bibliography and a glossary file
a linear document which cites a variety of texts, with links to the original sources where appropriate
two parallel strands, with the possibility of jumping to and fro between them using hypertext links (for example to represent opposed arguments in a debate)
summary information, with expanded versions of each item accessed via links
a web of multiply interconnected fragments.

In looking at a hypertext document, the user will normally be equipped with a pointing device, probably a mouse. When the user identifies a word or phrase which promises to connect to another chunk of text, the user clicks (or in some systems, double-clicks) on the item and the linked text is revealed. A wide range of methods is employed to deliver this simple result. For example, how is it indicated to the user that a given text item is one which the user must interact with in order to traverse a link? How does the new text appear in relation to the original? Are there facilities for the user to return instantly to the original text? In HTML, neither publisher nor user currently has much control over how these things are done.

Virtual documents

Bush speculated that if the physicality of individual documents could be done away with, advantages would include:

an unlimited number of instances of the same' document
virtual books: instead of documents residing in the place where they originally belonged, parts of texts could be compiled at need into new volumes'
trails: the very process of building a hypertext creating new links between existing document segments would be a new form of authorship, and these trails would have intellectual value of their own, usefully exchanged between scholars and perhaps the subject of commercial transaction.

Some problems with hypertext

Bush saw traditional document structures as inimical to research, and to constructive thought. However, in proposing other structures, he did not anticipate the problems which have tended to arise in practice. Of these the most prominent is the lost in hyperspace' problem, where users become disorientated as they traverse one link after another and are unable to get the best value from these apparently powerful structures.
Hypertextual documents are useful only if the structure...

adequately reflects the user's information requirements
is clear and simple enough for users to be able to find their way through
the content lends itself to being delivered in chunks'.

Anyone who is familiar with electronic mail will be aware of its tendency to abolish a sense of spatial separation, rather as telephony does. However, at least with e-mail, each document is a discrete unit arriving from a particular geographical location. The interlinking of hypertexts goes further, abolishing even this sense of place, since intimately connected parts of a document may be stored on opposite sides of the world. This lack of a sense of location raises issues for authors and document designers. Without knowing where a document is coming from' literally and metaphorically it is sometimes difficult to know how to read it.

There is a danger of making a maze of links, in which users travel around and around, rediscovering parts of the document they have already seen, unable to find the information they require. Documents can also become excessively fragmented.

Hypertextual interconnections can be advantageous, but there may be value in providing linear means of accessing those linked elements as well. For example, experiments have been done with guides' or agents' which attempt to lead the user through hypertext and hypermedia documents, describing points of interest or importance along the way.

The 'homeopathic fallacy'

Bush strongly held the view that documents should be organised like the mind'. Leaving aside the fact that controversy rages as to what the mind is like and how information is structured there, there is still a problem. To make a structure, even one which users can clearly perceive, is not to succeed in putting that structure into the user's mind. The educational task may be transformed, but it is not miraculously overcome. See Hammond et al, listed in Resources (p108)

Structures do not solve the problem of enabling the user to understand by mysteriously planting that structure in the user's mind. Users must still make an effort, requiring motivation and intelligence, to make the author's structures (or something like them) their own.

Contents

Graphics Multimedia Virtual Environments Visualisation Contents