Common nouns | A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | 123..
Proper nouns A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z| HOME

Arts of the text: Literature, Verse, Programming...

Calliope, by Albert Ducrocq, the first automatic poem generator (1953).

Last revised 8/29/2014. Return to Major concepts. See also language, poetry, storytelling, book, typography, prosody.

All the digital writers we know of are listed in our Index, with direct link on their Diccan's item. We concentrate here on conceptual considerations and the state of the art. The historical development of the field is described, period after period, in our history of digital art.

1. Combinatorial creation
2. Grammatical generation
Annex : programming
3. Storytelling
4. Games and interaction
5. The graphic chain and multimedia
6. Distribution and reading issues, transmedia
7. (not properly digital art but...) : Computer use for Literature study and analysis.


Text is digital by nature, word processing ubiquitous. But writing as a distinctly digital activity remains quite rare. The most digital are generative, which applies to litetrature, poetry and possibly live performance. Here we can distinguish the generation of sentences, using generative grammars or combination of text fragments. And we have also, at a more global level, the storytelling.

For each structure, art will reside in a proper and balanced combination of
- random and order. See ~[Leech] and our Laval paper Uncanny peak.
- formal features of the text and nature of the content.
Then, digital art in application of rules. Basically the spell checkers...

Text may be artistically used in a limited number of ways:
- Literature uses customary ways of speaking and writing, searching art both in style (something like application of restricting rules on sentences building) and content.
Not "artistic" but generative. Journalism.

Poetry submits itself to more formal rules, mainly rhyme and metrics. Note that "Electronic Literature" includes prose and poetry (see ELO).

Literature, as such, has not much to show on the digital table, but the works of James Joyce. He is considered a forerunner of serialism, in his way to use language, pushed at its maximum in Ulysses (1922). But everybody, including himself, agrees to recognize that this work is practically illegible, whereas it is a must for the shelves of any good library. Perhaps an exploration of the successive schools of poetry in this time could bring pertinent insights.

Hypertext did not open a really new field to literature or poetry.

Multimedia and transmedia are global forms of art, heavily depending on digital tools, but de facto using text in traditional ways, and so not creating a properly digital way of writing.

Literature, as properly a digital art, is a small world. Globally, indeed, books, press and radio amount to some 12 billion euros in the French economy, and nearly 200 000 jobs.
Digital tools are of course widely used all along of the publication chain. But they are not explicitly a source of artistic creation. Literature remains created by authors with their minds and hands (on a keyboard, of course).

Digital literature is more commonly called electronic, as proven by the name of   the main ad hoc association Electronic Literature Organization (ELO) . This activity is perhaps easier in the American-English world, where universities host "digital humanities" (see {Hayles] )departments, more oriented toward creativity that the French "Lettres" .

Communities are active also in several countries, and particularly in France, where Philip Bootz organized in September 2013 a full week of events, and has put online a vast review of the domain. Our base lists a good hundreds of writers which can be considered digital. But you won't find their works at booksellers, even at shops specialized in poetry. And forget the possibility of a Goncourt prize or a celebrated bestseller. Digital art is here, even more than in plastic arts or cinema, a practice of scholars and fans. Then, what do they do?

- On the political opposition side, we can quote Hans Haacke, a conceptual writer and anti-fascism activist.
- The graphic aspects of text may be artistically built, with calligraphy and typography. The sound aspects are dealt with prosody.

1. Combinatorial creation, or mounting from fragments

The basic form of properly digital literature is generative. Philip Bootz [Bootz] makes the distinction between to modes:
- "combinatorial": pre-built text fragments are assembled according to algorithms, most of the time simple pseudo-random,
- "automatic", building texts from a dictionary, syntax rules and stylistic habits of an author (or the preferences and taste of a creator).

The combinatorial form has pre-digital origins, notably the Oulipo group and its set of protocols for combinatorial text generation (mostly, but there are attempts in drawing, painting, theater). Later, William Burroughs is famous for his "cuts".

These techniques are still in use. In 2013, for instance, Magali Desbazeille used them on the derisive mode about digital art critics.

Stefanie Strickland is today one of the  stars in this realm. She works  by combination, using mainly a deep and long familiarity with the grand authors. For instance, Sea and Spar, which she created with Nick Montfort, is a "poetry generator which defines a space of language populated by a number of stanzas comparable to the number of fish in the sea, around 225 trillion. Each stanza is indicated by two coordinates, as with latitude and longitude. They range from 0 : 0 to 14992383 : 14992383.". Click on [34] and you can navigate in these space by longitudes and latitudes.
This combinatorial aspect can be graphically expressed, for instance Textrain by Camille Utterbach. It's an interactive installation. An image of the spectator, projected on a wall, intercepts a rain of characters. Proper gestures let letters combine into words and phrases.

The mounting may be visible by the reader. It is generally hidden in Oulipo-like productions. On the contrary, it can be emphasised, with artistic use of typographic and page setting effects, as in Only Revolutions by Mark Danielewski.

But this "oulipian" model has radical limits. If you rely only on random to cut out fragments then to assemble them, you get no more than chaos. You may have interesting results if you start with the construction of a corpus of fragments, sufficiently coherent in structure and semantics. We think that, to make new progress, the selection and assembly should be done according to a sufficiently elaborate set of rules and criteria.


Vannevar Bush : as soon as 1945, a vision of the hyper-communicating world of 21th century.

2. Grammatical generation

Several thinkers write about the formalisms in art, like Max Bense (poetry and aesthetic theory). During the 1950's Albert Ducrocq and his Calliope binary generator of poetry or the computer generated sound poems of Gysin Brion.

Jean-Pierre Balpe has written many books on the subject, and his software has written a lot ot text, including an opera. He is still head of the movement on the early 2010's. But his breakthrough in the 1980's showed quite rapidly the limits of the genre. For instance in imitating the style of Flaubert. He attained sufficient quality to deceive experts about this author: one of them, reading such a Balpe's automated productions, commented: "It's certainly Flaubert's, but I don't see from which work". But this way seems rather forlorn for the time being.

Grammatical generation, from a codified and standard set of meaningless (or weakling meaning) bitsets is a major avenue of generative art, though coding takes it farther.

Linguistic structures are here at home, with their generative grammars. An extensive review of the domain is given by [Bootz].  Here, the rules of assembly are clearly given by the dictionary and grammar... but only the structural laws and some schematic patterns (subject, predicate, complement...). Another set of rules can be derived from a formalization of an author's style (typically, Balpe for Flaubert, musical pastiches of Mozart, or generators of images à la Mondrian). The rest is let to random or to "manual" operations by the author.

Oriental languages make a complex and synesthetic use of grammar, since ideograms can have meaning both directly (for instance 中 for middle) and by definitions. Calligraphy and Haiku poems draw on this duality, and some attempts have been done to use it in digital works.

The generative field is cultivated by Rafael Perez-y-Perez (poem generators), Abilio Cavalheiro with his Sintex tool and even an art critic generator, by Eric Maillet. J.C. Carpenter generates fiction.

See : Computational Linguistics and Intelligent Text Processing by Alexander Gelbukh. LNCS 7817. Springer 2013. A philosophical thought "... philosophy and literature live in the same generative space, even though it is finally limited. Their performative means are identical: words alignment, syntactic modes, punctuation (a subtle resource). It's true as well for the nursery rhymes as for a Kant's critique, for a pulp novel as for Plato's Phedon. They are acts of language. The idea, cherished by Nietsche or Valéry, that abstract thinking could be danced, is an allegorical affectedness. Everything is in the wording, the intelligible statement". George Steiner, dans Poésie et pensée, Gallimard 2011 (our translation).


See some notes about programmming.
It seems difficult to go deeper than the grammatical assembly. But you can code grammar, and possibly generate irregularities, which are essential in poetry, as comments [Leech] for instance

. Programming by itself may be considered an art, since a program is a text, which can be aesthetically appreciated independently of its functions. That's certainly true in principle. But this kind of beauty can be appreciated only by other experienced programmers.

3. Storytelling and higher levels

See also the item storytellng and a chapter in the transmedia notice.

Narration Vs. Database

At still higher levels, we find a database/narration opposition which is quite similar to the mounting/grammatical seen above. The distinction is made notably by Lev Manovich [Manovich 2001]. It is commented, from a literary standpoint by [Hayles 2012]. It
- corresponds to different philosophies: the "grand narration" is "modern", the database (seen as a collection of short narrations) is postmodern. We find the same opposition, applied to psychology of the otakus in [Azuma],
- can be extended to any media and art form,
- can be extended into a wider scale, from the most linear sampling of events in time to multidimensional data structures, and many intermediate levels.


The narration is probably the most ancient form of literature. Small narrations are then integrated into grand narrations. The Bible is the most typical of this, with an integration of several ancient traditions into the official canons.
Principles for narration can be found in Aristotle's Poetics (around 350 B.C.) A lot on Wikipedia , but in many other Internet sites. In the pre-digital mood, Aristotle offers here a forerunner of "structural" analysis and rules for making (Greek poieien). Full text on line.

In a more scientific modes "concordances" are indexes from words into biblical texts. For example Handkonkordanz zum Griechischen Neuen Testament by Alfred Schmöller (Wurttt. Bibel Anstalt, Stuttgart 17th edition, c. 1950). Official doctrinal texts of the Catholic church have been indexed for instance in Denzinger's Enchiridion Symbolorum (Herder, Barcelona,1957).

Text from data, robot jounalism.

In 1999, Grégory Chatonsky exposed The Forest I (2009), a wall hung work of art, where Nasdaq financial data are transformed, real time, to a forest of pixels and synhetic voices. The artist dicourse sounded as an alert. Today, robot journalism is a reality.

Just type "robot journalism" in your favorite browser, and you get documents by millions. At present, it is more a pragmatic search for cost reduction than properly art. But the rest of this paper about writing tells clearly that narrative abilities and style will progress. Robots are a real threat for the journalist profession, yet deeply weakened by the take off of Internet Would it be only for economic reasons: the browsers are not artists, they just find the documents for you. But they collect the advertisement income, which yesterday (in the 20th century...) paid for the professional writers.

Some examples.

- Search for information:
- send robots and drones in difficult, dangerous or otherwise unaccessible places (Planète robots, Sept-Oct 2014, with gathers several sources on robot journalism).
. and of course browse the web and build your own databases and archives.
. build theories from the "Big data" and statatistics. (Le Monde 3/7/2014).

- Write papers:
. Narrative Science sells software and claims in its home page banner :"With spreadsheets, you have to calculate. With visualizations, you have to interpret. With narratives, all you have to do is read."
. Stats Monkey" does it for finance and sport.
. News At Seven (a project) automatically generates a virtual news show. Totally autonomous, it collects, parses, edits and organizes news stories and then passes the formatted content to artificial TV anchors for presentation. Using the resources present on the web, the system goes beyond the straight text of the news stories to also retrieve relevant images and blogs with commentary on the topics to be presented."
- Quakebot automatically generates a message when an earthquake is detected.
- Nearly 10% of the tweetos are robots (Le Monde, 8/14/2014).

- Detect lies: The Whashington Post Truth teller says you when politicians assertions are false.
- Compute the importance of a news or paper, where and when to broadcast it: Visual revenue tells you. The product is in use by several chains.
- Censor litigious contents: Crunch base does it (and a lot of other tasks).

- Make a synthesis: Yahoo News Digest prepares something like a "press review".

Database literature

The "database" here is not the collection of elementary facts or figures that you can find, for instance, in a spreadsheet file. It is a collection of short stories or poems, an anthology for instance, a book of lectures, an encyclopedia, etc. Each part has meaning by itself, and there is no meaning in the order or in the transitions from one piece to the following.

In the 1980's, strong hopes were raised by the Hypercard software on Macintosh computers. John Cayley invested a lot on this tool... he is not the only one to have lost data and creative activity when the product is dropped by Apple.

The Google's answer to a query can be considered as a form of "database" text construction.

The readers have more freedom in the use of "databases", they are more "engaged" with great narrations. But the border is not sharp. The Bible may be considered as a grand narration, from the creation to the Israeli state. But it is also a collection of many texts, each one consistent by itself, and sometimes with no direct link to the adjacent "books". A "database" use of the Bible is possible due to the segmentation of the texts into chapters and verses. A preacher, for instance, will dig into the Bible to feed his sermon. A less official use of the Bible, nearer to pagan oracle practice than to religious orthodoxy, takes a page at random to answer a question.

That leads us to a new attitude towards text: not something definitively "given" by an author and "received" by a reader, but a work calling for interaction wih an active "spectactor".

4. Games and interaction

Digital tools gives the reader new powers, in respect of the work, extending what could be done with paper. Copy, mix, machine reading. Extracts. Look on Internet for critics, commentaries.

Some beautiful applications of interaction, combining image and text, have opened new ways, which remain essentially limited experiences and installations

Particularly pleasant, we feel, are the poetic interactions of the spectator with rains of letter, by Camille Utterbach and also Robert Cahen. Nitika Pashenkov, with humor, animates interactive characters in his Alphabot.

Edouard Lussan goes up to interactive comic strips.

Camille Utterbach: Text rain (1999).

A major form or textual interaction was given as soon as the 60's by the Eliza machine. It was not art, and has been highly disputed, even by its author Joseph Weisenbaum. See [Wardrip-Fruin, 2009].

Hypertext affords non-linear navigation into sufficiently fragmented text. Some comments, both critical and poetical, are given in Mark Amerika's  Metadata. [2].

Storytelling can become interactive. Hypertext has had high times in the 1980's CD-Roms. See [Murray, 2008].

5. The graphic chain and multimedia

Immaterial by nature, the text needs nevertheless to take a material form to exist and be perceived by us. There are two forms: typographic (on paper or screen) and vocal. Both may be the occasion of artistic efforts.

Typography and calligraphy

The fundamental bases were laid by with Donald Knuth's algorithmic researches. New fonts creation or enhancements are regularly proposed, for instance by François Boltana, but he uses the computer as a tool, on the screen, like a painter on Photoshop.

Page setting is now accessible to anybody through the HTML codes, completed with CSS styles and many tools. Easy page setting and content management are free on the web with CMS (content management systems) such as Wordpress.

The new technologies have progressively impacted the whole graphic chain, down to the printing presses themselves. That reduces the costs and increases quality. It is difficult in the 21st century to remember how poor press and publishing in general were limited in their presentations: few fonts, time consuming page setting, high cost of corrections. It's evident for instance when you open an issue of L'illustration, which was the luxury French magazine from 1880 (circa) to the 2d world war?

Kurt Campbell associates text and sculpture.

As for calligraphy, Kaalam, does not consider himself digital, but we do not resist to the beauty of his performances and recordings.

Graphics are now comparatively easy to combine with texts, and many artists take the bait. For instance the collective work Scriptura et Caetera (1999), by Marie Belisle, André Savard and Michel Du_Bois, A set of poems and short texts which change according to geometrical shapes or Sale Temps (1997) by Frank Dufour, Gilles Armanetti, and Jacky Chiffot, a kind of photographic compositing. And also Komninos Zervos.

But the most busy track combines text and graphics on different ways. Camille Cholain develops a form of derisive hypertextual poetry on the web. Loss Pequeño Glazier animates poetry on a screen. Wade Guyton prints very large canvasses with few letters.
Kasper Sonne. inserts text in his paintings, and Frank Laroze writes the multimedia way.

See book. Other references on the subject:
< Esthétique de la ponctuation, by Isabelle Serça. Gallimard 2012. Not digital.
< La raison graphique ou la domestication de la pensée sauvage. by Jacques Goody. Editions de Minuit, 1979.

Prosody and performance

On the sound side, oral text reading by computer has been available at low cost for decades. But it never reached a quality of recognition and expression sufficient to use them as a proper form of art. It may just be a complement in ordinary life devices (telephone services, train stations announcements). In fact, our ears and more generally hearing system is very sensitive to weaknesses in speech, and rapidly teased if not exasperated by repeated faults and by excessive repetition itself.

Even digitally generated, text used in live performance must be read by human voices. Such presentations were a major part in the Cherchez le Texte week in Paris (2013).  Thas goes along with cooperative works, as for instance Poème percutant, played by Philippe Bootz, Nicolas Bauffe, Claude Moreau and Jean-Pierre Moreau.

Many "poets" combine poetry with other devices/medias: performance for Jerome Fletcher and Cécile Portier, games for : Jason Nelson, directed sound for Sandrine Deumier

Some references
< The prosody: Claude Maillard's Théâtre de l’écriture (1986), a polyphonic text, for three voices plus a metronome.

Audiovisual, cross-media, multimedia

Talking movies were a first emergence of image and sound combined. In the 1960's the term "audiovisual" was used, mainly for didactic and communication applicatiosn. Then came multimedia and crossmedia (1995).

Digital techniques add new ways to the combination of text with sound, image, now the web, and the performance presentations of texts.

Theater, songs and opera use text as component of live performance (possibly recorded), and we deal with them in separate chapters. And, finally, text can simply play its part in multimedia and transmedia.


Some references:
< Writing for Multimedia and the Web, Third Edition: A Practical Guide to Content Development for Interactive Media , byTimothy Garrand

Writing about writing...

< Exploring Web and Multimedia Writing: The Art of Words in a Visual World, by Juliet Davis (Seems out of stock).

6. Distribution and reading issues, transmedia

With the traditional text media, the author's and reader's role were rather clearly cut, as a producer working for a customer, somehow. Digital technologies change the picture at least along four directions:

1. The work itself has a form of autonomous life, and particular is able to interact more or less actively with the reader (see above).

2. The reader is offered multiple ways of using the text, from simple viewing on a screen to printing it... and of course (at least for some decades), to buying the book. [Hayles] analyses different modes of reading. Entertainment, information, close reading, hyper-reading machine reading. These modes are partly independent of the work itself. You can fragment it, combine it with other works (and medias) and do your own way through.

3. Authors and readers are playing in a networked world, where everything can be copy/pasted, forwarded, debated over; then the authors and/or the publishers have to bother with the complex circulation in this network,

4. In this networked world, "robots" (algorithms, distribution systems) take a more and more central place, increasing the "distance" between author and reader (though sometimes reducing it). More and more, authors have to write "for the robots" even more than for their final readers. To say worse: we are foreseeing a world where robots write for robots. Humans are still here, perhaps, but each year farther from the central activity and even creation.

Transmedia, in the most general sense of the term, brings here another ways of reading. Katherine Hayles comments it in depth with his book How we think [20], which would perhaps better be titled "How we read". She comments notably on the difference between "close reading" (stay long and concentrated on one and only text) "hyper reading" (read by surfing around), stressing the values and the necessity of both practices and "machine reading" (using software to study the texts).

For further information and comments, see our dedicated notice to transmedia.

7. Computer use for Literature study and analysis.

(Should be developed). A firts reference : Informatique et Bible Maredsous.

Some references

< L'évolution du traitement de texte, par Michel Volle. 2015. Online.
< History of Computer Art. Online. Interesting mainly on the early history of hypertext literature. Text, Speech, and Dialogue by Ivan Habernal, Václav Matoušek. Lecture Notes in Computer Science, subseries: Lecture Notes in Artificial Intelligence. Volume 8082. Springer 2013.
< The website NT2. Laboratoire Nouvelles technologies, Nouvelles textualités (New technologies, new textualities) of Uqam (University of Québec in Montréal).
< Ecrivains sans papier. A 5 pages article by Christine Ferniot. Télérama 6/6/2012.
< Filtrage sémantique. De l'annotation à la navigation textuelle by J.L. Minel (ed.). Hermès Lavoisier 2009.
< L'écriture des médias informatisés. Espaces de pratiques. by Cécile Tardy and Yves Jeanneret eds. Hermès-Lavoisier 2007.
< Internet et écriture (2005, 2009) An on line PhD thesis by Eric Guichard. Interesting notes about digital compositing and Powerpoint software (from a human and sociological standpoint, not technical).
< Ergonomie du logiciel et design web, par Jean-François Nogier. Dunod 2003 (2d ed)
- Le texte au coeur de la nébuleuse informatique. by Pierre Berger Le Monde informatique, 1994. A synthesis, softly futurologist, seen from 2010.
< L'écrivain public et l'ordinateur by Jean-Jacques Salomon and André Lebeau. Hachette 1988
< La civilisation de l'écriture, by Roger Druet and Herman Grégoire. Fayard 1976. (Expensive book).
- [Laufer] and [Scavetta] have retraced, in the 1970's the art applications of typewriters.
< Le pendule de Foucault. Grasset 1990. (Italian edition 1970). With pages about text processing.
< L'oeuvre ouverte, by Umberto Eco, Chantal Roux de Besizieux and André Boucourechliev. Seuil 1965.