Skip to content

Your browser is no longer supported. Please upgrade your browser to improve your experience.

Digital Literature Pioneers: Judy Malloy on ‘narrabases’

In 1986 Judy Malloy published her novella Uncle Roger online as a narrative database or ‘narrabase’. In the first part of two-part interview she speaks about her artistic and digital influences, e-literature today and the Silicon Valley scene of the 1980s.

We’ve linked to simple English definitions of some of the more esoteric references.

 Read part two of our interview with Judy Malloy here.

You released Uncle Roger in 1986, describing it as a “narrabase”, can you tell us a bit about both the work and this term?

I began writing the words for Uncle Roger and programming an authoring system to access them in August of 1986. My vision was to create a computer-mediated novella in which the reader individually recreates a fictional environment by continually searching and retrieving narrative information.

In this work of fiction that emerged from a vision of the possibilities of computer-mediated narrative, I intertwined elements of magic realism with Silicon Valley culture, semiconductor industry lore, and early word processing. I wanted the work itself to reflect the transitioning of computer culture. Thus the content, like the choice of hacker’s BASIC as a programming language, was set in Silicon Valley in the era of semiconductor “chip” wars — where the race for who could make the fastest chip heralded the era of personal computer culture.

Based on a creative use of links, (originally called keywords from the database technology and algorithms that informed this work) the first file of Uncle Roger, “A Party in Woodside”, invites the reader to follow chains of links through the narrative. Like the experience of attending a party, each reader emerges with an individual picture. Note that although the first person narrator, Jenny, is the central character of Uncle Roger, in computer interface practice, she also functions as an “agent” whose presence personifies the experience.

Originally authored for command line computer platforms, the three files of Uncle Roger are created with hundreds of lexias – screen sized units of text that can either stand by themselves or be combined with other lexias – to create a coherent narrative. Uncle Roger was structured with three files because I thought that issues of time and place in non-sequential literature would work better if the narrative data was decisively delimited. Remember that at the time, there weren’t models for what I wanted to create.

The three “files” of Uncle Roger are:

“A Party in Woodside”

During a long, mostly sleepless night after a party in a Silicon Valley bedroom community is remembered fitfully, interspersed with dreams. As at any party, the reader meets some people but not others, observes some events but not others.  In this experimental narrative, the compressed time and place of a party was also important in the build-up of coherent detail.

“The Blue Notebook”

In Silicon Valley, things do not happen simply and clearly. In “The Blue Notebook”, parallel yet intertwining narratives advance the story in sometimes-conflicting ways. Reflecting the increasing complexity of the narrator’s life, “The Blue Notebook” remixes time and place/reality and magic reality. As the narrator herself observes: “The things I wrote in the blue notebook didn’t happen in exactly the way I wrote them.”


To break the formal search and retrieval structure of lexias in the first two files, for the more diffuse experience of the third file, “Terminals”, I wrote a program that – simulating the unpredictable ways we remember – randomly displays a new lexia every time the reader presses return. “Terminals” takes place shortly after the time when computers had replaced typewriters in corporate offices. The reading of the story continually changes as the frustrating, challenging, boring, or productive office experience is randomly intertwined with remembered events in the narrator’s life.

The whole Uncle Roger is a “narrabase”, or narrative database. Perhaps it is time to return to the future and revisit this approach.


In 1998 you said that Uncle Roger is “not a means of providing alternative plot turns and endings.” This is even today a complex approach to digital literature that considers computers might enhance a reader’s experience of story, beyond simply allowing the reader to choose different endings. Were there many other people thinking this way at the time of writing? What/ who were your influences? 

Italo Calvino, Marcel Proust, Dorothy Richardson, and Laurence Sterne.

And then there was my experience with designing databases and my contingent idea that there was no reason that a database always had to be confined to information that was not narrative or poetic. I was familiar with the vast databases of the Library of Congress, having worked on the Union Catalog, and I envisioned that large, full text narrative and/or poetic databases could be constructed.

The ideas for the algorithms of Uncle Roger and for the use of small but complete units of narrative text that could be combined with other units of text to create a meaningful work of literature occurred in this way: in 1969 I was project head for the creation of an early library database at an aerospace company in Boulder, Colorado.  To do this, I studied FORTRAN and Systems Analysis. From this knowledge came the vision to create a non-sequential literature.

But at that time computers were often screen-less, room-filling machines.  So in 1976 when I began trying to create a databased work of literature, I chose to make one of a kind artists’ book. At first I used trays of catalog cards and push button electromechanical books to tell non-sequential stories with words and pictures.  Then in 1986, a rare event (in anyone’s life) occurred. Everything came together.

It began with the Apple II; it began with Art Com Electronic Network (ACEN). The catalyst was my old friend, San Francisco-based curator Carl Loeffler. In 1986, Carl and Fred Truck were setting up a radical online artspace on the WELL, and Carl called me up and said. “You have got to go online.”

So, I had some ideas I had been developing about experimental literary forms; I knew how to program; and unexpectedly I had an audience and an online publisher on Art Com Electronic Network!

I did not know about Ted Nelson’s work until a few years after I had created Uncle Roger. I did know about the work of information specialists, such as Henriette Davidson Avram and Ralph H. Parker that enabled the creation of large systems of searchable computerized library catalogs. Occasionally I wonder why the heroes of library automation receive so little Digital Humanities credit.

I was familiar with the Gysin-Burroughs “cut-up” method, but although the cut-up technique was of interest, I wasn’t fragmenting a whole, I was writing hundreds of lexias, each one written separately in such a way that it could either stand by itself or be combined with other lexias. Nevertheless, ideas build on each other, so I think what they did is important to the field of electronic literature as a whole.

The same is true of Adventure games – what is now termed Interactive Fiction. In retrospect, although my vision was to build on the artists books I had started in 1976, the existence of early Interactive Fiction made my return to computer programming seem more feasible.  However, there is no treasure, no mystery, no maze in Uncle Roger. Rather there is a series of small scenes. Indeed, scene-based Renaissance comedy was influential not only in the method but also in the “jig” content that sometimes occurs in Uncle Roger.

As well as the world’s first “narrabase” Uncle Roger was one of the world’s earliest online novels [published on the ACEN bb boards or forums]. In 1991 you described the benefits of publishing online: “Online systems facilitate immediate publication, are compatible with any computer with a modem, and integrate the artist with the audience.”  How has the concept of the online novel been transformed by web 2.0?

Beginning with the now long ago story of Art Com Electronic Network. ACEN was an incredibly forward-looking early social media platform. Under the leadership of alternative art space director, editor and publisher, Carl Loeffler and in collaboration with system operator, Fred Truck, ACEN was seminal in the creation of a comprehensive pre-web online environment that included the publication of online journals; the interactive online distribution of text-based artworks and electronic literature; and a conferencing system for discussion, with topics such as “Software as Art”.

The menu Fred Truck created with UNIX shell scripts was elegant, and ACEN had access to The WELL‘s server to enable our work to run interactively. Nowadays, most systems operators don’t (for good reasons) allow this level of access, so it is difficult to convey how this worked seven or eight years before the World Wide Web became prevalent.  It should also be noted that Loeffler and La Mamelle/Art Com were known and respected in the art world. Kathy Acker and Taylor Meade were among many contemporary experimental writers who performed/read at Com/La Mamelle. And the first work published on ACEN was John Cage’s First Meeting of the Satie Society. But Carl’s vision also encompassed pop conceptual art and video art, and one of his heroes was Andy Warhol.

There was a lot of “buzz” and interactive reader interest when Uncle Roger appeared on the WELL in 1986. Although the audience on the Web is much, much larger, even in an online literary magazine such as the Iowa Web Review, on the Web, works of literature exist in an environment of slick commercial sites.

However, with the advent of a more interactive web 2.0, readers are beginning (or returning) to explore electronic literature, while at the same time, thanks in part to organizations promoting electronic literature, we are creating a body of known “public literature”.

I am optimistic about the future of electronic literature, and I’m looking forward to co-teaching Electronic Literature Lineage, Theory and Contemporary Practice at Princeton (with Cliff Wulfman) this fall. We are now in an era when interest in the possibilities for screen-based literature is returning – both for readers and for writers!

Could you give us some insight into the database / query language you were using at the time? How did it influence the content/ story itself?

From the very beginning, I wrote Uncle Roger with a “keyword field” for each lexia, and as I developed the story, the keywords and key-phrases became a part of the process.

So, when I told the story in serial fashion online on the BBS in 1986, all I had to do was include the keyfield with each lexia, and readers could put the work into their own database software. This strategy worked and allowed for any query language that the reader’s database software supported. But ultimately it did not create an aesthetically pleasing screen, nor (at that time) could I create an online version or a stand-alone version using the available database software packages. Thus I wrote a UNIX shell script version for publication on the online ACEN interactive database menu, and I used my BASIC version for the stand-alone version.

Yes, the process was complex, (and time consuming) but remember this was an era of rapidly changing software and platforms.  Any developer, even a writer of electronic literature, had to continually “translate” his or her work.

Except for the web version, which in some cases substitutes graphic icons for keywords, (an interface device I used in 1981 for the card catalog The Woodpile) access to “A Party in Woodside” and “The Blue Notebook” was through database query from a list of keywords. The reader could either search by one keyword/key-phrase or use the Boolean operator “and” to combine two keywords – for example “men in tan suits” and “food”.

There was not an option to search for any word in Uncle Roger, although Cathy Marshall and I did include this option when we authored Forward Anywhere some years later. The differences between reader experience in a keyword/link controlled search, such as I used for Uncle Roger, and allowing the reader to search for any word are of interest, as is the role of authorial control in electronic literature.

Have you worked with a different database since then? And if yes how have you found it?

In 1988, I retooled the program I created for “Terminals” (Uncle Roger: file 3) to create the original version of its name was Penelope. Although I love returning to Uncle Roger  – for its energy, its aura of lowbrow Renaissance comedy, its core existence in my body of work, its redolence of Silicon Valley in a certain era, and for the memories of the incredible experience of working on ACEN that are somehow a part of Uncle Roger -– in retrospect, Penelope more elegantly fulfilled my original vision. The iPad version of Penelope that Eastgate will be releasing soon is intuitive and wonderful!

In the creation of Penelope, once again there were issues of authorial and reader control, for instance, whether a complex interface is desirable or whether it works better to conceal all traces of the authoring system. In my work, this is often determined by the content and by the character of the narrator. (Compare The Roar of Destiny – where the narrator works on the Internet – to Dorothy Abrona McCrae, where the narrator is a painter from an older generation – to its name was Penelope, where the narrator is a conceptual photographer)

In 2003, I based A Party at Silver Beach -on the web interface for “The Blue Notebook”.  And I brought back Jenny, the narrator of Uncle Roger to tell this story of another party.  Although Silver Beach seems very different from Uncle Roger, if you think of the icons as keywords there are similarities. However, Uncle Roger was a more classic use of the database structure.  A lot more could be said about the differences and similarities, but onward…..

And then in 2009, for the main interface page of Where Every Luminous Landscape (part one of Paths of Memory and Painting) I wrote the lexias in 8 tracks. The tracks are related to each other in that they reveal the narrative in different ways. I could have chosen to let the reader follow these tracks by searching, as I did in Uncle Roger, but instead, *all on one top page* I exposed the first lexia of each of the 8 tracks. The reader can follow any one path by continuing to click on the lexias in that part of the array. Additionally, each lexia is timed to replace itself after a certain amount of time; so each track in what is essentially a database (one could even say a relational database but there might be some argument on this) that exists in an environment of shifting narrative. Where Every Luminous Landscape is a complex work, but I think it is wonderful in its way.

Judy Malloy is an artist and expert in electronic literature, currently Visiting Lecturer, Program in American Studies; Council of Humanities, at Princeton University. You can find out more about Judy’s work here and  Uncle Roger here.


Back to Archive