The contents of this page may seem familiar to you. It draws heavily on the ideas of Conklin [1987] which he wrote down in his 'Hypertext: An introduction and survey'. I consider this an essential work for a thorough examination of hypertext systems.
Already in 1945(!), Vannevar Bush came up with the idea of the Memex, a system that would contain scientific literature, as well as photographic images and personal notes. The user should be able to link different entries. The concept of the never build Memex influenced later hypertext pioneers, like Nelson [1980]. The following fragment describes the Memex. It makes you think of a computer applying a desktop interface metaphor.
[..] A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory. It consists of a desk, and while it can presumably be operated from a distance, it is primarily the piece of furniture at which he works. On the top are slanting translucent screens, on which material can be projected for convenient reading. There is a keyboard, and sets of buttons and levers.
Nelson once defined hypertext, also known as non-sequential text or non-linear text, as
a combination of natural language text with the computer's capacity for interactive branching, or dynamic display [...] of a non-linear text [...] which cannot be printed conveniently on a conventional page.
Please note that intertextual references are not new. The importance of hypertext is simply that references are supported by a computer.
In our opinion, The COOPerator may be considered a hypertext tool. The following requirements, given by Conklin, who states that a hypertext system is a union of three powerful and well known metaphors, prove our point:
In case of The COOPerator, the 'database' at the client-level is the directed connected graph. The host-application maps this graph into a relational database system. Extensive support for operations on links, by means of a consistent windows interface, has been provided (e.g., links and nodes are created, traversed, and retrieved in a transparent manner).
It is probably best to think of hypertext as a computer-based medium to support thinking and communication. The thinking process does not build new ideas one at a time, starting with nothing and turning out each idea as a finished pearl. Thinking seems rather to proceed on several fronts at one, developing and rejecting ideas at different levels and on different points in parallel, each idea depending on and contributing to the others (Conklin).
Traditional flat text restricts one to writing and reading in a mostly linear order. Of course there are tricks for signalling side steps in the flow of thought: comments, footnotes, intersectional and bibliographic references, and sidebars all allow the author to inform the reader about related thoughts. According to Conklin these are rough tools at best. They often do not provide the degree of precision or the speed and convenience of access - it may be necessary to consult indices and turn pages - that one would like. Hypertext allows and encourages a writer to make references, and allows a reader to make his own decisions about what links to follow and in what order.
In The COOPerator project, we have extended the hypertext model to the more general concept of hypermedia. In hypermedia the elements that are linked together in the network could be text, graphics, audio recordings, animation, digitised speech, and full motion video. Similar to Nelson's definition of hypertext, hypermedia or a non-sequential, multimedial document is
a combination of multiple media elements with the computer's capacity for interactive branching or dynamic display (the playing of sounds inclusive) of a non-linear document which cannot be printed conveniently on conventional paper.
The last part of the definition is true in two different ways: not only are students unable to print the entire document in a linear fashion on conventional paper, including annotations, alternative document parts or the underlying discussion, it will also be very difficult to obtain tangible copies from multimedial document parts like sounds and video fragments! In the remainder of this paragraph, the term 'hypermedia' will be used.
A hypermedia system should require no more than a few keystrokes or mouse clicks from its user to follow a link. The interface should enable links to act like buttons in order to transport the reader quickly and easily to a new place in the hyperdocument. It is essential that the system responds quickly to the referencing requests as the reader often does not know if he wants to pursue a link reference until he has had a quick look at the referenced node.
Although the essence of the hypermedia model is its machine-supported linking, the nodes contribute significantly to defining the operations that a hypermedia system can perform. Most users of hypermedia systems favour using nodes which express a single concept or idea. Hence, these nodes are much smaller than traditional files. The hypermedia model invites a writer to modularise his ideas into units. He may have to face some difficult decisions about how to separate ideas into suitable pieces of information.
A hypermedia environment offers new possibilities for accessing large or complex information sources. A linear document can only be read in the order in which it is composed. The essential advantage of non-linear documents is the ability to organise objects in different ways, depending on different points of view or different demands.
Conklin sums up a number of advantages of hypermedia which more or less apply to the COOPerator:
Apparently there are two different classes of problems related to hypermedia: problems with the current implementations and problems that seem to be characteristic to the hypermedia model.
The problems in the first class are mostly interface problems. These are 'merely' technical shortcomings, expected to be solved somewhere in time. Typical dilemmas in this category include delays in the display of referenced material, lack of browsers or deficiencies in existing browsers, and the like.
The other category of shortcomings offers, according to a few influential authors in the field (e.g., Conklin and Nielsen [1990]), more challenging problems that may in fact ultimately limit the usefulness of hypermedia: disorientation and cognitive overhead.
The risk of disorientation while navigating the information space (also known as getting lost in hyperspace) is one of the major usability problems with hypermedia systems. For example: studies by Nielsen showed that 56 percent of the readers of a document that was composed in one of the most popular commercial hypermedia systems (HyperCard), agreed fully or at least partly with the statement: I was often confused about 'where I was'.
Of course, a disorientation problem also occurs in traditional linear documents. But in a linear document a reader only has two options: he can search for the desired object further on or earlier in the document. The hypermedia model offers more freedom, more dimensions in which the reader can move, and hence a greater potential for him to become lost or disoriented. So along with the power to organise information in a much more complex way, students have to face the problem of having to keep track of where they are in the network and, even more difficult, how to get to some other place they know (or just think) exists in the graph. In a large network information can easily become hard to find or even forgotten! Thus, hypermedia environments can enforce a rather harsh information hiding.
According to Conklin there is another fundamental problem with using hypertext: cognitive overhead. It may be difficult for group members to become accustomed to the additional mental overhead required to create and keep track of links. In general: the additional effort and concentration necessary to maintain several tasks or trails at one time may be experienced as a burden.
This problem does not only occur in the process of constructing hyperdocuments, it also frequently happens while browsing the hyperspace. A student will often be presented with a large number of choices about which links to follow and which to leave alone. At the moment he encounters a link, how does he decide if following the side path is worth the distraction? This must sound familiar to anyone who has ever used the Microsoft Windows® on-line help facility which is mostly loaded with numerous hyperlinks.
Next to that: When the computer display is small (which is very likely to be the case for our target computer configuration), group members can only see a very small part of the information at any one time. This means that they can very easily loose track of how the fragment they are currently watching is related to the immediately preceding or following media object, since that fragment will often be invisible at the time. Nielsen calls this a problem with context-in-the-small as opposed to the before mentioned problem regarding context-in-the-large, the equivalent to 'getting lost in hyperspace'.
Sjoerd Michels, Tilburg, The Netherlands.