return to index
Duck Song
(Text and image in [on-line] narrative)
In 1997 this essay was entitled `The uses of Text'.
A version was presented as a paper on May 9th at INFOG 97
Another version appears in Cantrill's Film Notes #85,86 June 1997
And yet another was included in British Film Institute, Vol 25 #3 Aug. 1999
For ease of reading the essay is designed to be printed
Updates and relationships to the work it describes are available [on-line]
[Voice] 61 3 94897905 [Email]

(CC) reserved Simon Pockley
  1. Summary
  2. Types of text
  3. Context
  4. Hypertext
  5. Book-like text
  6. Morsels of text
  7. Encoding text
  8. A universe of text
  9. Image as text
  10. Space for text
  11. Scholarly text
  12. Screens for text
  13. Participants text
  14. Poetics of text
  15. Notes & External References
1: Summary

Writing for the World Wide Web involves the use of different forms of text. Some of these forms draw on long standing literary conventions, others are unique to the web, where many of the distinctions between image and text have become blurred. The Flight of Ducks takes the textual form of a journey through a landscape and turns it into a contextual universe where hypertext paths can be taken through a datascape. These paths form stories. Narratives that can dip into their paper bound origins or plunge into the poetics of the screen space. In this space they are composed into shimmering pixilated displays where image and text are inseparable and anyone can participate.

2: Types of text
To write by fragments: the fragments are then 
so many stones on the perimeter of a circle: 
I spread myself around: my whole little universe; 
at the centre, what?

                           Roland Barthes 1

It is not a new idea that the structure of a work should reflect the nature of its content; that form conveys meaning. For Aristotle it was part of the resonance that made a drama 'work'. Form and meaning were inseparable for Marshall McLuhan, and Jacques Derrida called the phenomenon a mise en abime (arrangement to the extreme). 2 Describing the use of hypertext, Greg Ulmer gives to this wholeness the rather awkward expression reflexive structuration, a means by which text shows what it is telling, does what it says, displays its own making, reflects its own action.

So it is with The Flight of Ducks where an old travel journal is used not only as the primary source for a larger journey but also as a model for how to use and write for this electronically networked medium of the World Wide Web.

If you are looking at a paper-based printout of this text, it should be noted that while it may be a convenient form of textual representation, it is not the only one. This is also a hypertext document 4 with doors to examples, elaborations, illustrations, notes and sources. More importantly, it has an electronic context without which much of the content of this text will be meaningless.

At a binary level, all representations of text are abstracted forms of electronic coding. There are at least seven different types of text used in The Flight of Ducks and in order to explain how they contribute to narrative structure it is useful to be able to distinguish between them. When seen from the perspective of their original purpose, they fall into three distinct (but often overlapping) categories:

Paper based text: handwritten in pencil, pen and by typewriter.

Screen based text: encoding tags, hypertext, email and metadata.

Printable text: book-like text, scholarly text.

These words form an encoded hypertext essay. They are given structure by following a line of thought which starts with the conventions of paper based (document-like) text and follows various types through to non-document, not-printable uses of text. In some cases this progression is revealed through the development of the site, in others it is not so sequentially convenient. Diversions into the poetics of the screen space will need explanatory groundwork. If this suggests some form of hierarchy of text, it is not intended.

3: Context
Stand on one of those seemingly boundless gibber plains, 
the horizon of the whole circle as unbroken 
as if you were far out on the ocean, 
stand there if you wish to know your own proportion in
the scale of the visible world.

        		Robert Croll (1937) 5

In Australian literature the travel journal has been as closely associated with the physical landscape as with an imaginary landscape of the mind. Interest in this form of writing is usually reserved for the picturesque or for the accounts of explorers whose heavily revised field journals were imaginatively re-shaped into epic struggles of survival in a hostile and alien land. It is not hard to trace the way the literature of exploration converges towards the centre of the continent. But from the outset, the quest for an inland sea was such a tantalising idea that it was inevitably accompanied by disappointment. The successive waves of rediscovery since the 1930s continue to show how imaginary destinations are not just projections of need but an essential part of the journey itself.

Over the last century the 'known' and mapped landscape has been peppered with new destinations. Natural features are now promoted by tourist operators as essential stops on a journey of self-discovery. But the projections remain, perhaps as they have always done, drawing people inland towards an inward journey with its perfect physical counterpart in the landscape.

There have been many published accounts of central Australian journeys since the 1930s. 6 Post exploration, the Aborigines themselves became the objects of discovery, initially, as anthropological objects and then as the remnant repositories of an eroded traditional culture. Many of these accounts were written by men with a connection to scientific expeditions. Most are nostalgic of a frontier that receded with the expansion and upgrade of roads during World War 11. Others, such as Mountford's Brown men and Red Sand (1948) try to revive the explorers experience by presenting the account as if it was a first sighting of the country and its inhabitants. Mountford chose to travel by camel and took photographs of Aborigines without clothes, even though the country had been criss-crossed by roads and settled for over 20 years. The iconography of the naked Aborigine persisted into the late 1960's. In film, it was replaced by the sounds of didgeredoo as an evocation of a mystic relationship with the land. However, most of these references simply tried to share and articulate the author's profound and poetic experience of the Centre. Together they lay the foundations for the most recent forms of this genre. Books such as Bruce Chatwin's Songlines (1987) and more recently Barry Hill's The Rock (1994) are, in fact, spiritual journeys assembled from the imagination, field notes, and previous accounts.

The 1933 field journal of my father, F.J.A Pockley, was written at the very end of the frontier period, when there were still isolated groups of Aborigines (known as wild blacks) who had yet to have white contact. He had been a member of a scientific expedition based at Hermannsburg Mission but out of boredom and a thirst for adventure, hired camels and travelled west along the MacDonnell Ranges to Mount Liebig with an Aboriginal guide, Hezekiel; animal and skull collector, Stanley Larnach; and artist, Arthur Murch.

Although my father was only twenty years old when he wrote this journal, he rewrote it several times before his death in 1990 and changed it considerably (mainly through elaboration). He never sought to have it published. Indeed, he insisted that it was not of sufficient interest (or quality) to warrant publication. At first glance, he was right. The original diary, though anchored in its time and now of some interest as a primary historical source, is, by itself, little more than a few sparse jottings. In fact, until work began on The Flight of Ducks , only the passage of time pulled the original text back from the edge of ephemera and oblivion.

As a narrative account of a journey into central Australia it underwent many of the revision processes that served to shape published accounts and shows narrative and structural characteristics peculiar to this form of writing. In Living in a New Country Paul Carter writes:

Explorer narratives are characteristically discontinuous. They also lack `plot'. On these two grounds alone they are of theoretical as well as historical interest: The means by which they advance the story are pertinent to understanding the mechanisms of fiction; they also suggest that the current narrative/non-narrative debate in historical circles is based on a false opposition. It may be that, under certain conditions, narrative can respect the discontinuity of historical action. (Paul Carter 1992)7

Even though it is not the journal of an explorer, Carter's insights into this type of writing lead us towards its significance as a model for a narrative use of hypertext.

4: Hypertext

The development of a re-centreable system based on hypertext should be seen in the larger context of post-structural theory where discontinuity provides a focus for another process of exploratory convergence. The literary theories of Jacques Derrida and Roland Barthes and the computational theories of Theodor Nelson and Andries van Dam, meet in their rejection of a primary and fixed centre of organisation of ideas. 8

For many of the Marxist semioticians who have dominated intellectual inquiry for the last thirty years, the lineal progression of the book represented all that was outmoded and tyrannical. From its chapters to its sentences and punctuation, the book displayed a fixed central line of authoritarian thought.

The anticipation of hypertext by these critical theorists and the reciprocal embodiment of these theories by its implementation on-line, is no less historically remarkable than the prescience of Vannevar Bush. In the mid-1930s, (roughly the same time as my father's journey and before the desktop computer) he had the idea of the memex. This machine was like a desk with translucent screens, levers and motors. It could be used for the rapid searching of records and had a capacity for contextual association where 'any item may be caused at will to select immediately and automatically another'. His description of how it was to be used is an eerily accurate description of the use of hypertext.

When the user is building a trail, he names it, inserts the name in his code book, and taps it out on his keyboard. Before him are the two items to be joined, projected onto adjacent viewing positions. At the bottom of each there are a number of blank code spaces, and a pointer is set to indicate one of these on each item. The user taps a single key, and the items are permanently joined. In each code space appears the code word. Out of view, but also in the code space, is inserted a set of dots for photocell viewing; and on each item these dots by their positions designate the index number of the other item. Thereafter, at any time, when one of these items is in view, the other can be instantly recalled merely by tapping a button below the corresponding code space. (Vannevar Bush 1945) 9

This cognitive trail, as a trace, a thread, or path through a databased landscape actually makes reading resemble writing. A path is recorded in such a way that a functional narrative is constructed out of a labyrinth of discontinuous possibilities.

When using the Internet, few of today's travellers in this datascape, take the trouble to record their path or to pay attention to the trace they leave unless they are actually engaged in the process of writing or are retracing this path. Certainly, in the physical world, travel accounts are usually retracings through memory at the end of the day or even at the end of the journey. In my father's case, time, memory, even perceptions themselves are often ordered by the action of recollection and shaped into a narrative that stretched through time to give structure to his life. 11

5: Book-like text

The idea of the path as an ordering mechanism has also been implicit in the development of theories of text processing and encoding since the 1980's. Many of these theories have been based on the idea that text is composed of an ordered hierarchy 12 of nested objects such as chapters, sections, paragraphs, extracts, lists and so on. They are called objects because they organise the text into natural units based on meaning or intention. The structure is hierarchical because these objects have a linear relationship to each other. Sentences, for example, exist within paragraphs. This is essentially a book-like view of text and it is reflected in the most commonly employed markup language, SGML (Standard Generalized Markup Language). 13

There are two book-like forms of the journal employed in The Flight of Ducks. Neither has anything to do with semiotic concerns or with paper based origins, but everything to do with portability. The first is a book-like version (Fig. 1.) of the edited hypertext (1933) journal with embedded photographs, the second, a book-like journal of the journey retracement, in 1996. The second acts as a postscript to the first so it seemed appropriate to retain the form.

screen shot of book-like page 
(13.347 kbytes)

Fig. 1 (book-like display for printing)

At first I scoffed at requests for a book-like version because I saw it as a poor substitute for the electronic hypertext. However, once I began planning to retrace the expedition, the need for a portable non-electronic copy made it obvious that such a version would have many advantages in the field. It would serve as a comprehensible book-like object to show people with no understanding of the World Wide Web. The ability to print out a book seemed like an interesting functional extension of a web site. A book-like form, when printed out in chapters and with headings, seems like an obvious and simple thing to do - it was not. Even in mid 1998 the printing of these hypertexts with page numbers has been a cumbersome process.

To date, the ease of text publication on paper afforded by the word processor is not matched by any such facility on-line. The complexities of page formatting encoded text with embedded images are manifold. Proprietary 'web editors' exist to perform the transport from paper/word processor to web but not in the opposite direction - from web to paper - particularly when the text and images have not been sourced from a word processed environment.

The text of the book-like version is a combination of all the variants and was only possible because these variants had already been combined in the hypertext version. As hypertext, these variants existed as fragments or smaller units (lexia) which had to be cut and pasted to form the larger units of each chapter. The breaks between chapters were not native to the original but based on links to paths and were largely determined by the paths of the hypertext version (I refer to this later).

A book-like form needed photographs integrated and collated with the text so that their locations and content could be checked during the1996 retracement. Embedding the photographs within the text was extremely time consuming because their screen sizes bore no relation to their physical sizes. This meant that they had no physical representation beyond the screen. Each of the 200 images, therefore, had to be re-sized for a paper space and placed so that when each page was printed the page break would not cause image enjambment (dividing the image between pages). I also wanted each page to be numbered. Bifurcations in the hypertext were generally resolved by footnotes that again complicated page layout and so were placed at the end of each chapter. The large byte size presence of the images meant that chapter divisions were also determined by the need for manageable file sizes requiring each chapter to be displayed on a separate screen. A division based on icon paths did not work for the camel path because the file sizes were simply too large, so this path was broken into four.

While I was able to complete this task successfully on my own equipment, the process was not a universal success because there are printers that make a mess of my formatting and make the book-like facility less than perfect. It was interesting that the process of making these book-like versions changed my view of the material. I had previously considered the bulk of the source material to be primarily text but once the images were embedded I found that I was looking at a picture book with textual explanation.

Electronically these book-like versions have qualities that paper based ones do not. To examine how these are used and to prepare the ground for a more detailed examination of the breaking up of text into discrete units, it is necessary to ask:

What is electronic text?

6: Morsels of text
You can't tie a knot with your tongue
that you can't undo with your teeth

                      Lila Stapleton (1979)14

Taken simply as a word, `text' carries from its derivation notions of texture and woven fabric. It is certainly more than a sequence of binary oppositions represented by the abstractions of 0's and 1's. While deconstructionists attempt to reduce all forms of text, even punctuation, to the level of distinctive signifiers or marks, pragmatic analysis seems to focus on a unit of text as a block of words (or sequence of images). These are referred to by Roland Bathes as a `lexia' but more commonly known as a 'text chunk'. Lexia can be identified by the extent to which they can be detached.

The first step of decomposition is the bite and the organ of this new philospheme, is the mouth, the mouth that bites, chews, tastes. . . the `mourceau,' the bit, piece, morsel, fragment; musical composition; snack, mouthful. This mourceau is always detached, as its name indicates and so you do not forget it, with the teeth, and these teeth, refer to quotation marks, brackets, parentheses: when language is cited the effect is that of releasing the grasp or hold of a controlling context". (George P. Landow, Hypertext) 15

Determining what constitutes a lexia is first and foremost a function of its useful purpose or perspective. In this example it is being used to illustrate functional characteristics of lexia in an electronic medium. Gregory Ulmer calls this citational graft 16 Theodor Nelson calls it transclusion. 17

7: Encoding text

Text encoding is the method by which these textual units are gift wrapped using markup so that they might be put into another context by rearrangement or association.

There are several encoding languages of varying degrees of complexity that have the ability to create hyper-media. The spectacular development of the World Wide Web has been largely based on the simplicity of HTML (HyperText Markup Language). To a large extent The Flight of Ducks is an encoding project based on the application of HTML.

At the time of writing HTML has a number of advantages:

HTML's primary disadvantage is that (at this stage) it will not support many of the more sophisticated analysis functions available in SGML from which HTML was derived.

It is also important to consider the role hypertext has played in the narrative development of The Flight of Ducks.

The Flight of Ducks began with a comparison of the original pencil version of the 1933 journal with that of a contemporary but incomplete copy in pen and a later extended type written version. My aim had been to produce little more than an annotated record of the variants. These editorial annotations (which took several years) are now irretrievably lost to the corrupted hard drive of an obsolete machine. As a work of textual scholarship it was probably misconceived. A detailed comparison of variants might be appropriate to the scholarship of major works of literature but this journal is not one of The Canterbury Tales 18 although it provides a fascinating example of the corrupting effects of memory.

In spite of the severe limitations of the medium, in 1995, the World Wide Web offered a number of advantages over what was then considered an appropriate computational outcome for this kind of material - CD-ROM. The first and primary advantage was that this computational work was no longer platform or software dependent and so there was less chance of losing it. Second, it made the material accessible by a wide audience regardless of its significance. Third, the markup language was so simple to learn that it could accommodate my lack of experience or skill in multimedia production. The praise and awards The Flight of Ducks has received proves that the site has found an audience 19 and shows the importance of simplicity and accessibility.

The Flight of Ducks has been an evolving documentary-like work drawing on various forms of the original journal. It shapes its stories into the larger narrative that, in many ways, is still unfolding. The addition of further journals, notes, sounds, photographs and related metadata (including this explanatory text) provide a context for the original journal. They also lend the text and the story, levels of transparency and resonance that could not have been anticipated and that now lead to a rich texture of meaning and significance.

8: A universe of text

Julia Flanders 20 (the textbase editor of The Women Writers Project) 21 in considering the new emphasis on what she calls electronic copia, asks, "How much data is enough?" When people feel impelled to provide everything, they may be trying to compensate for `an anxiety about their ability to represent the real (as against the virtual) world'. She sees a calling for the use of images to substantiate what would otherwise seem to be a radically untrustworthy source of information.

Similarly, the goal of creating not an edition but an archive -- of providing all the source materials necessary for the reader to form his or her own analysis -- is surely rooted partly in the impulse to transport an entire textual universe into the new medium, to give the electronic edition a kind of self-sufficiency that can substitute for whatever physical reality it seems to have lost. (Julia Flanders 1997) 22

There are three important points here, all revolving around the process of `transport'. The first is that the `textual universe', Julia Flanders refers to, is probably a natural characteristic of any new medium where new works lack the context they have when they grow out of a tradition. This could well be part of the process (to paraphrase McCluhan) where new media makes old media content. It is certainly not always seen as inappropriate or excessive. In fact, it is Theodor Nelson's idealised docuverse and Derrida's vast assemblage. It leads to the second point, concerning 'self-sufficiency' which grows out of a primary need for context. The ability of hypertext to create associations makes it a contextual mechanism. Literary theories abound with questions concerning the degree to which a work needs to be embedded in contextual data in order to have meaning: Do we need to know anything about the author? Do we need to know anything about the culture? Is it possible for a work to stand alone? Third, her sense of loss of physical reality is probably related to her focus on encoding extant text. Foxing and rust marks infer the physical qualities missing in electronic texts that have become so abstracted as to have no physical existence at all beyond their shimmering, mutable, pixilated representations.

9: Image of text

For Jay David Bolter23 this is part of a larger tension between the word and the image where images are seen as more natural signs that require a different form of engagement. The substantiation Julia Flanders calls for, is important to The Flight of Ducks because it ties the work to its literary context and to the reality of a real journey. From my own perspective it has been interesting to see how the importance of representing the original journal has receded as the contextual world of the work has grown. Attempts to annotate the original pencil version with either the pen or the type written text, have been abandoned in favour of the more powerful narrative that has emerged as a result of placing the original journal into the larger context of extrapolation, recollection and retracement. This has created a line of hypertext that acts as a kind of narrative spine running through the work.

Without this spine The Flight of Ducks would probably have no shape and dissolve into a series of tracks leading nowhere. Similarly, without any indication of the breadth of the site or the placement of this spine The Flight of Ducks could be perceived as a collection of jottings. The structure of the datascape mirrors the physical landscape through which the expedition travels.


Fig. 2. Datascape - The eastern spine of the MacDonnell Ranges

The original pencil version runs parallel to this narrative spine like the ridges of the MacDonnell Ranges (Fig. 2.). It not only functions to substantiate the hypertext but also gives it a context. The persistent presence of this version draws attention to the textual differences. For the close reader it poses a problem: why are these texts different? The answer is a study of time and memory drawing from the original a narrative structure layered through time as the story unfolds. At a meta-level the breadth of the work is formalised in the data represented in the main (combing screen (overview). On a narrative level, a linear search or the pursuit of a line of inquiry creates the paths. Consider the following collection of lexia:

screen from Page 42 of original

Fig. 3. Text 1 - From the original (press for full page)

As we now had a good reserve of meat I set off for a prominent hill to the south, on foot. Something about it caught my eye when we were favoured by the panoramic view and I had seen it again from a ridge a couple of days before. I doubted if I could make it as, at my best guess, it was about forty miles away and I knew water would be a problem. I decided to walk as much as possible at night and took a full water bag, a rifle and one blanket, rolled as a swag. I had shot off all my film for the camera round the native camp and it was too heavy to carry anyway. I knew that there would be enough moon and that I could steer by Pavo. The others were all against it, but I felt that it was something I had to do, largely because I was half scared and had had it on my mind for days, with a growing fascination.

Text 2 - from the hypertext spine (c.1970)

I was on the shady side and the ground looked awesomely remote when I looked down, for the first and only time. I ate the apple slowly in order to prolong my stay, but forced myself to edge out onto the face again in the end, rather acutely aware of diminishing courage. I climbed more slowly now taking infinite care with foot and hand holds which seemed to get steadily harder to find and more awkwardly placed. The pitch was just short of vertical, but had where the surface was more glossy and smoother and perhaps rather darker in colour. I began to stop climbing and to cling to the face, rather than rest, fighting off a looming panic.

Text 3 - from the 1976 return

Looking north it seemed unlikely that he had indeed be able to see The Olgas from Mount Peculiar. I felt that in some way I was looking at The Olgas with a view to climbing. What a pointless endeavour that would be - the son trying to match the father - for what? While thinking about the significance of all this I recalled a sonnet I had written to him in 1979 when Susan became pregnant with Emily. Interesting the way imagery enters the subconscious:

Text 4 - from the 1996 retracing

This is a narrative that spans over 60 years, implicit in its structure, but requiring more than just a superficial reading. It can be traced through the work by filling in the gaps between the lexia. It is significant because the formation of narrative is independent of the text. Indeed, it is still unfolding because a close reading of the original shows that the journey on foot was probably imaginary. The meaning of this is another matter entirely.

Because substantiation relies on the noise (stains and creases etc) of each of the handwritten pages for its verisimilitude, the digital representation of this text (Fig. 3.) is enhanced by a slight sharpening and increase of contrast. This also has a practical benefit of making the pages more legible than they are in the original.

Attachment to the physical medium is further emphasised by using a miniature version of each page to represent the link between the two forms of text. This miniature draws on the iconography and positional conventions of the illumination in medieval manuscripts where the text flows from a miniature illustration. Such a metaphorical resonance (another practical application) also provides a source for an appropriate low-resolution precursive image that allows time for the high resolution, image to arrive.

<a href="FOD0718.html"><img src="images/pncil56.gif" lowsrc="images/page.gif" alt="Original journal written in pencil page 56 (199 Kbytes)" width="565" height="708" border="2"></a>

As a text unit, the original pencil journal is so bound to its paper medium that the digital representation of each page shares most of the characteristics of the surrogate photograph (file type, size, specification etc). The distinctions that do exist, emphasize the powerful role of representation (as opposed to content) in determining lexia.

In The Electronic Word, Richard Lanham 24 makes a distinction between two kinds of textual engagement: that of `looking through' and `looking at'. Looking through means being transported by the meaning of the text and looking at means examining its representation. Usually we oscillate between the two. Looking at the text of the original journal makes the order of the pages sequentially neutral. More importantly, as lexia, they can have quite different meanings depending on which side of the oscillation is attracting attention. As substantiation (as in the previous example) the lexia are saying 'Yes, these pages are evidence that these events did happen' because I can see what are obviously old pieces of paper with handwriting. When being looked through (or read in sequence) the reader has to conclude (as I did - and it came as a shock) that 'No, this did not happen.'

Before looking at the difficulties inherent in having parallel journals it is necessary to look at another determinant of lexia most commonly ignored or denied by web writers and that is the role of the screen space.

10: Space for text

One of the strongest limiting factors to the reduction of size of computer devices is the size of the screen. Regardless of resolution, size is limited by our ability to discern type. The threshold of legibility is somewhere between 9 and 12 points. Once this threshold is reached, less data can be shown and more interactions with the screen will be needed, ironically, requiring more screen space.

While large computer screens are becoming more common, the platform independence of the World Wide Web means that client side screen size cannot be assumed. Good web writing practice, therefore, means that a lowest common screen size of 640 x 480 pixels is usually adopted as the writing space (Fig. 4.).

screen shot of writing space 
(7.960 kbytes)

Fig. 4. Screen integrity in the 640 x 480 pixel writing space

This space effects lexia in diverse ways. The most obvious is that large bodies of text are not easy to send down-line as discrete lexia and therefore have to be broken up into smaller file sizes. Less obvious but I think, more important is a desirable information arrangement within the screen space that I refer to as screen integrity. In this space all accessible information is visible in one instance without having to scroll up, down, or sideways to find hidden text or images. Integrity implies an unimpaired wholeness and relates to a comfortable appropriateness of form. Certainly, there are bodies of text, such as lists and this text, where scrolling is acceptable. But when acknowledged, spatial considerations further refine the notion of lexia.

Any random access of the World Wide Web will show what little value or awareness is given to screen integrity. I suspect this is because most users have yet to free themselves from the spatial demands of the paper page with its more flexible aspect ratio. Evidence for this is immediately apparent in the use of the term 'home page' or just 'page' when these users are referring to screens or display space.

11: Scholarly text

Since 1995, the proliferation and distribution of on-line academic publications (still referred to as 'papers') on the web have led to the development of conventions around a form of text that may be transitional. Designed to be printed out has come to mean that the text is a single file and instructions to print will result in the entire text being printed. Text in this form is usually only broken up if the file size is too large to be downloaded efficiently as a single file (tolerance depends on transfer rates). The body of the text is usually preceded by links to the host site. An abstract or brief description is followed by internally referenced hypertext contents in the form of a list of descriptive section headings. Links throughout the text are either internally linked footnotes that appear at the end, or external references to other sites, or both. The best also include email contact details.

Often screens that don't work are the most instructive. It is a curious phenomenon that the worst (difficult to use) examples often belong to authors of hypertext theory such as George Landow. 25 On the one hand, want they their texts to be screen based but, on the other, retain most of the paper based conventions. Typically, it is difficult to get any sense of the breadth of these texts. They meander through what appears to be a labyrinth with broken links and encoding errors until one loses faith in their currency. They are invariably broken into a series of small files for no apparent reason with little or no screen integrity and with obscure navigational aids.

The hypertexts contained in this research report have an encoding architecture that makes them appear [off-line] to be fairly simple documents. [On-line] they are complex organisms. Short of creating two separate documents the outcome is a compromise between the two forms. 26

12: Screens of text

What happens when [off-line] writing is adapted to be read on-line?

The spinal hypertext of the edited journal was initially broken into lexia largely determined by the 640 x 480 screen-aspect ratio. The basic unit was the paragraph. This was not always available as a natural break either because it ended prematurely or because it was too big for the screen. Where this occurred, new breaks were inserted at the end of sentences. Beyond these constraints text is left free to float so that any variation in aspect ratio on the client side retains the integrity of the screen by simply re-shaping the text to fit the space without disturbing the position of the navigation icons.

A larger set of text units was indicated by the iconography of the navigation links. These links allow the participant to travel forwards and backwards through the text:

First, the idea of progress was implicit in the images/icons of transport: train truck camel etc. Each indicated a lexia by an icon drawn from the photographs. The only exception to this was the time spent in Alice Springs where I used the image of a windmill which was not only common to the town but has a Quixotic reference (similar heritage). Movements between the paths when using these icons are seamless apart from the changing icon (Fig. 5.). A more formal reference to lexia was created by dividing the main combing screen into separate paths.

screen shot showing changed icons 
(5,193 kbytes)

Fig. 5. Seamless movement with changing icons

Second, the progress of time is emphasised by diary-like entries on every screen. When these dates are vague, (time became so extended during elaboration that dates were lost) the month + cont. is used. Following the date is a brief summary of the content of the text. This serves as a quick form of reference to the passage of narrative. It is drawn from the conventions of the eighteenth century picaresque novel. It also acts as metadata to the journal entry where it is reproduced in the <title> tag and the combing screens. The encoding source code (Fig. 6) for the screen above (Fig. 4) is as follows. The display text is left dark:

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "">
<title> Diary 1933: ducksong </title>
<link rel=schema.dc href="">
<meta name="DC.Identifier" content="">
<meta name="DC.Relation.IsPartOf" Content="The Flight of Ducks">
<meta name="DC.Relation.IsReferencedBy" scheme="DCTERMS.URI" 0259.html">
<meta name="DC.Relation.Requires" content="Browser supporting CSS or better" >
<meta name="DC.Creator" content="Pockley, John">
<meta name="DC.Source" scheme="DCTERMS.URI" 0870.html">
<meta name="DC.Type" content="compound/mixed">
<meta name="DC.Type" content="journal">
<meta name="DC.Format" content="text/html">
<meta name="DC.Format" content="image/gif">
<meta name="DC.Rights" scheme="DCTERMS.URI" 0310.html">
<meta name="DC.Keywords" content="The Flight of Ducks, documentary, history, journal, Haasts Bluff, 1933, Aborigines, sing-song, duck flying away song, corroboree, possum songs, black man,">
<meta name="DC.Description" lang="en" content="Hypertext journal entry describing the camel expedition from Hermannsburg to Mount Liebig in January 1933. This entry describes John Pockley's impressions on hearing the duck flying away song.">
<meta name="DC.Coverage" content="Central Australia; 1933">
<meta name="DC.Language" content="en">
<meta name="x-Sequence" content="89">
<meta name="DC.Date.Available" scheme="WTN8601" content="1995-05-05">

<link rel=schema.AC href="">
<meta name="" content="Pockley, Simon">
<meta name="" content="">
<meta name="AC.activity" content="modified">
<meta name="" content="1999-07-26">
<meta name="AC.rights" content="FOD0310.html">


<div class="header">
<a href="FOD0004.html">
<img src="images/Lds3.gif" width="40" height="45" border="0"
alt="return to index" />
</div class="header">

<div class="epigraph">
28th January 1933 cont. - the duck flying away song
</div class="epigraph">


<a href="FOD0725.html">
<img hspace="6" src="images/page.gif" align="left" width="38" height="48" border="0"
alt="original journal reference">

We heard amongst many others the duck flying away song. It is impossible to describe, but it was rather frighteningly effective. Rhythm was paramount and the pitch tended to rise to a crescendo while the pace quickened in most of the verses. Forms were inverted and the character of the black man's voice gave it all something indescribable, at any rate we all knew that the ducks were resting on the water, were surprised, and took off clumsily then flew off and away but returned with a swoop at terrific speed to disappear into freedom again and peace. </p>

<div class="navigation">
<a href="FOD0033.html">
<img src="images/Lcml.gif" align="left" width="80" height="80" alt="back" border="0">

<a href="FOD0034.html">
<img src="images/Rcml.gif" align="right" width="80" height="80" alt="forward" border="0">
</div class="navigation">

<div class="footer" align="center">
(CC) reserved Simon Pockley Feb 1995 The Flight of Ducks


<email><a href="mailto:"></email>
<img hspace="8" src="images/mail.gif"
align="top" width="35" height="15" border="0"
alt="send message" /">
</div class="footer">


Fig. 6. (Source code) Title and summary within metadata and markup.

The hypertext spine contains links to notes, images, further elaborations and subsequent journeys. All of which would be difficult to access so readily in a printed text. The very act of moving rapidly backwards and forwards between these links and different lexia runs counter to the inherent hierarchical structure of book-like text. Each lexia is capable of functioning simultaneously in different roles. For example, a participant might have searched the combing screen for `half-castes' and found the photograph of the children at the Old Telegraph Station. The icon next to the photograph takes her to the text describing this photograph. The text is now functioning to explain the photograph rather than being part of a linear narrative. She might then move to the note about these orphanages or move into the hypertext to see the context of this information. The important point is that she might arrive at this screen from many directions and therefore from many perspectives. The text is free to form into different lexia and meanings.

Apart from a couple of loose pages in the original version, one containing drawings and the other a map, there are a number of navigational difficulties inherent in having parallel journals with different representations of textual units. The first and most obvious is that they become unsynchronised. Variants and elaborations create quite different sized lexia. This is most apparent when returning from the original to the hypertext because the hypertext is made up of smaller lexia. Three or more fit each page of the original pencil version and enjambment is frequent. Sometimes there is simply no correspondence because of a lengthy passage of elaboration. My first solution was to use duplicate screens with different return addresses, but this became clumsy and was eventually abandoned in favour of small overlapping sequences. The Hyperwave browser is being developed to the point where multiple addresses are possible for a single screen.

Multiple layers of dynamic transparent text may be a better solution. The scripting necessary to make this a viable alternative would, at this stage, lead to a platform dependence outside the aims of this project.

13: Participants text

A form of text suffering the least abstraction from its origin is the participant's text. This text arrives by email as ASCII text. The body (without personal salutations) is simply cut and pasted on top of the previous message in chronological order. Remnants of the original email include a date and a return email address. Usually these follow email etiquette in that they are short and to the point. It has not been difficult to incorporate them as lexia without editing. The editing which does take place is usually for reasons of propriety.

An interesting extension of this has been the posting of questions sometimes quite beyond the scope of the project. In some cases answers have arrived (after some time) from other participants. I have diligently sent these replies to the questioners only to find that they have already received them directly and conversations have begun outside my sphere of influence.

In this way the site seems to be acting as a kind of campfire around which people are gathering. Several participants have sent me some of their photographs from central Australia as attachments.

This generosity provides an insight into the need for further extensions of The Flight of Ducks. By accommodating other people's interpretations of the material and even allowing them to alter the original material, their passages can become another experience or perspective. The transclusion 27envisaged by Ted Nelson is not yet available in a workable form, but the browsers are moving toward this facility. Difficulties with the Ted Nelson's Hyperwave browser development are at present delaying proper testing.

14: Poetics of Screen Text
The new writing might seek fluidity and reuse, rather than foundations and definitive position. It might provide paths that bring us to read a given lexia more than once. The form of the text is rhythmic, looping on itself in patterns and layers that gradually accrete meaning, just as the passage of time and events in one's lifetime. (Kolb 1996 28)

As a way of experimenting with the more active form of participation that this development promises, I began visiting the site as myself as a participant and writing my own path. Over time the content of the screens of this path have become more and more simple and linear. The form of text that seems to work best is almost poetic, forming itself into couplets, triplets and quatrains.

The origins of this began several years ago in another part of the site, where I had written a sonnet to describe my engagement with the project. My father was a prolific sonnet writer and our most intimate communications were by sonnet. But the fourteen lines of the sonnet form did not suit the screen aspect ratio without using a font size too small to read.

linear path 
(7,093 bytes)

Fig. 7. [Press image] for current version of this display

I broke the sonnet into octet, quatrain and couplet and sent these to separate screens so that they would suggest a rhythmic progression. In some places this progression is enhanced through the use of rhythmic imagery. One version of this (Fig. 7) shows how the image comes in to view with a pulse fitting the first words:

(This version may be accessed through a [Reload] of the NLA archive -July 1997) The image moved almost into focus and then moved out again with a movement representing the meaning of the octet itself. This has now been simplified even further. The image has broken free of the text and the text has been reduced to couplets. Progression is made sometimes through linking images, sometimes through linking text. In some screens there is no text at all. One path loops through a huge cycle of one hundred Aboriginal faces with unknown names (extracted from the photographs). The rigour of this poetic form is further emphasised and celebrated by presenting the text as plain ASCII using preformatted tags in markup, so that, unlike the hypertext spine, the text does not float. The hyperlinks gain depth and texture by drawing from the collection of archival content. This tightly drawn linear thread sits quite comfortably into the contextual universe described by Julia Flanders (see note 20). It seems to have a resonance that reflects the way we use the World Wide Web itself - following a line/idea through a larger context - a datascape.

These lines of memory run straight into the hypertext spine and merge with it - are engulfed by it. Curiously, they have grown in reverse to its direction of travel, in the same way that rain water flashing is installed or fish travel upstream to spawn. The tiny lexia of quatrains, triplets and couplets are constantly evolving and spawning others as this line contaminates the interstices of structured data.

It brings this line of thought full circle to the question in the epigraph by Roland Barthes at the beginning:

at the centre, what?  

If there is an answer, then it lies somewhere between the accounts of these journeys to the Centre and the screens and traces provided by the participants in this site. Their movements (and mine) through this datascape provide a shifting centre of perspective - a viewpoint. I tend to present this representation as having conventions closer to conversation than to traditional and current forms of historical narrative. However, the level of discussion and appreciation (below) suggest that we might be on the threshold of learning how to sing in this space.

Current writing is often personal, painfully so. But you've managed to personalise a story in a way that retains its multiple voices - yours (modulated differently as you move from sonnet to monograph), your father's (over the span of half a century), and the various embedded documents and reader's comments. The result isn't conversational, but composed. Each voice hits a separate note with the layering becoming almost musical.

And it may be closer to mythical than musical. All I remember from reading Levi-Strauss is that a myth is not a single story. It consists of all the retellings, the bastardised versions, musings, adumbrations, typos, jumbled memories, and even the anthropological commentary. Its the whole package. In your web site this layering begins to surface, becoming conscious. And the metaphors it uses are no longer passive but wired.

(Scott Thybony - email 1997, Dec 10.)

15: Notes & External References
  1. Roland Barthes, Roland Barthes, trans. by Richard Howard. New York: Hill and Wang, 1977.

  2. Landow, George P. [1992] Other Convergences: Intertextuality, Multivocality, and De-centeredness , Johns Hopkins University Press, P.10. [Online]. Available: [1998, April 28].

  3. Ulmer, Gregory L. [1985] Applied Grammatology: Post(e)-Pedagogy from Jacques Derrida to Joseph Beuys. Johns Hopkins University Press,

  4. The URL for this paper: Singing the Duck (

  5. R.H.Croll, Wide Horizons, Wanderings in Central Australia, (Angus and Robinson, 1937) P.1

  6. These are just the books in my own collection of this genre (in chronological order). The titles and subtitles are interesting in themselves. There are, of course, many more - particularly coffee-table books with photographs.

  7. Paul Carter, Living in a New Country (Faber & Faber 1992) P.11

    In this examination of exploration narratives Carter identifies thematic devices unique to the narratives of Australian explorers. He observes how (other than through attack or threat of attack) the Aborigines have little narrative impact, in the sense that their presence is either ignored or given the same status as features in the landscape. He also points out that 'Australian explorer journals differ from classic accounts of colonial exploration elsewhere in that they do not culminate in major discoveries.'

  8. Nelson, Ted. [No date] Xanadu [Online]. Available: [1998, May 28].

  9. Bush, Vannevar [No date] As we May Think ASCII text version: First published in The Atlantic Monthly, 1945 (section 7).[Online]. Available: [1998, May 28].

  10. In fact, our trail is recorded by the browser automatically.

  11. A broad outline is as follows:

    As a 20 year old F.J.A. Pockley may or may not have left the expedition and walked from Mount Liebig to Mount Peculiar by himself in 1933. From the summit he imagined he could see The Olgas across Lake Amadeus. He spent the next 43 years obsessed by The Olgas, but was encumbered by a busy and successful career as an opthalmic surgeon. The onset of a genetic desease which caused his father (also an opthalmic surgeon) to lose effective vision made him decide to build an inner life. He taught himself Ancient Greek and set out to learn the Iliad off by heart. In 1976 he approached The Olgas from the west and climbed central Mount Olga with the intention of singing the Iliad back to where he had been as a 20 year old. The climb was terrifying. All imaginative flight was lost to the practicality of surviving the descent and to the larger realisation of limitless human folly.

  12. Renear, Durand, and Mylonas [1993] Refining our Notion of What Text Really Is: The Problem of Overlapping Hierarchies [Online]. Available:

    [Online]. Available: [1998, May 28].

    An excellent paper which looks at some of the problems for SGML inherent in a book-like approach.

  13. Pockley, Simon. [1998] Killing The Duck to Keep the Quack [Online]. Available: [1998, May 28].

  14. Lila Stapleton was a neighbour and witness to my wedding in Mudgee, N.S.W. These words were the consolation she gave to my wife immediately after we we married. I have found them useful in numerous contexts and they seem extremely appropriate here.

  15. Landow, George P. Hypertext, [Online]. Available: [1998, April 28].

    The Passage quoted is from:

    Other Convergences: Intertextuality, Multivocality, and De-centeredness, [Online]. Available: [1998, April 28].

  16. Gregory Ulmer, Grammatology Hypertext, section 4

  17. Nelson, Ted. [1993] The Xanadu Ideal [Online]. Available: [1998, April 28].

  18. Blake, N.F. [1995] The Canterbury Tales [Online]. Available: [1998, April 28].

  19. Access statistics at Cinemedia show that non-robotic downloads of screens from The Flight of Ducks averaged approximately 5,000 per week during 1997.

    In late 1996 The Flight of Ducks shared an award with the DOS based CD-ROM game The Dame was Loaded which had a budget of over $1 million. Even in such a short time the advance of hardware and software have left this game unplayable.

  20. Julia Flanders, Editorial Methodology and the Electronic Text [Online]. Available: [1998, April 28].

  21. Flanders, Julia. [1998] The Victorian Women Writers Project [Online]. Available: [1998, April 28].

    This work aims to produce highly accurate transcriptions of literary works by British women writers of the late 19th century, encoded using the Standard Generalized Markup Language (SGML).

  22. Flanders, Julia. [1996] The Role of the Electronic Edition [Online]. Available: [1998, April 28].

    Naughton, Russell. [1998] Adventures in CyberSound [Online]. Available: [1998, May 28].

    A large work exhibiting this tendancy towards the creation of a contextual universe around the subject of wireless and radio. This work just keeps growing as it gathers to itself a comprehensive collection of material about the history of broadcasting at a time in which the medium is moving toward narrowcasting. The methodology by which this site absorbs material is not unlike Ted Nelson's transclusion.

  23. Bolter, J.[1996] Degrees of Freedom [Online]. Available: [1998, April 28].

  24. Richard A. Lanham, The Electronic Word: Democracy, Technology, and the Arts, The University of Chicago Press 1993

  25. Landow, George P. [1992] Reading and Writing in a Hypertext Environment, [Online]. Available: [1998, April 28].

  26. In fact, for the purposes of complying with Higher Degrees Committee guidleines these Hypertexts have had to be reformatted as `Word' documents in order to have page numbers etc.

  27. Nelson, Ted. The Xanadu Ideal [Online]. Available: [1998, April 28].

  28. Kolb, David. Socrates in the Labyrinth P. 341

  29. Squier, Joseph. [No date] Life With Father [Online]. Available: Life_With_Father.html [1998, April 28].

See also: