• Is a certain transformation designed as an artwork or artistic project of its own? If so, would it then be conceived to belong to the poetic program that the pre-text is representing? Or could we instead observe an opposition, perhaps as travesty or parody? Could this new work convincingly claim to be innovative or would it rather mean a trivialization of the pre-text?

• Is the recreation meant to be a translation or transmedialization? In these cases, we could discuss how appropriate it was created as compared with the pre-text.

• Do we have to assess the transformation not as an artistic expression on its own but as a creative reading and interpretation of the pre-text in a new technical environment – be it for research or for didactical mediation?

We will now discuss some examples from Portugal and Germany against this theoretical background. We offer two perspectives: an “internal” one reflects the production aesthetics of recreations (Torres) and an “external” one discusses projects from a reception aesthetic point of view (Block).

 

Poesia Encontrada by António Aragão

António Aragão (1924-) was the organizer of many publications of experimental poetry in Portugal. He has developed many pioneering experiences with electrography, photocopy, collage, or image-writing, and his work, very poorly published and mostly unavailable, is nonetheless acknowledged by all poets of the group as very innovative and influencing.

In trying to explain his collage-poems, Aragão writes that these cases of “found poetry” come entirely from the newspapers, claiming that the “malleability of the expression allows for several readings.”⁷ He further explains that “We supply one reading (…) [and] we leave also to the reader the possibility to make other readings, in other words, to do, to a certain extent, his/her poem.”⁸ This openness of the original text, as well as its projected fragmentation and accepted indeterminacy, can be well developed in digital media, given that cybertexts can be dynamic, open and randomly generated.

Aragão also predicted that this type of poetry, based as it is on everyday reality itself, “transforms and recreates reality in one another, an implausible one”. And he additionally maintains that “poetry has always varied its expression. (…) The incessant search for new and different expressions, which pervades all the arts, is tremendously provoked by the necessity of knowing and of giving (all) realities of the human spirit. Undoubtedly, the conquest of other expressions, instead of destroying poetry, is its way of passing through time, and it is its continuous and inexhaustible power of metamorphosis that confers it with validity and presence in the world.”

We believe that a work with these characteristics – impermanence, openness, indeterminacy – can be developed in a new form of expression, in a new medium, thus representing an interpretation of the pre-text in a new technical environment, in this case, the Internet. In truth, these found poems are based on the incompleteness and the indeterminacy of the text that supports them, because the text would differ according to the changes in the newspapers, in addition to the fact that the spatialization of the text in the collage also creates different readings and combinations.

For the recreation of these found poems we have used Jared Tarbel's Actionscript code of "The Emotion Fractal", "a recursive space filling algorithm"⁹ which places an arbitrarily sized word anywhere within a given rectangular area. In this way , we expect to put forward medial self-reference, in that the visual expressiveness of language is used in random typographic events that occur on the page. In effect, the words and expressions taken from online newspapers are spatialized on the screen in a random manner, imitating the collage aesthetics that seems to be implied in Aragão’s project.

On the other hand, as Tarbel explains, the effect of this code is “a region of space completely filled with increasingly smaller type. (…) The actual word placed is randomly determined, taken from a predefined list of English words.” In this case, however, the text, instead of being a copy of Aragão’s original collages, i.e., instead of being a reproduction of the pre-text, is dynamically pulled from rss feeds of online newspapers on the web. The result of this php script is an ever-changing, dynamic collage, different every time it is accessed: both visually and verbally. At the same time, this recreation reflects the poetological category of networking: the poem is live updated from web servers, being its text dependent on those changes.

Máquina de Emaranhar Paisagens by Herberto Helder & Pedro Barbosa

 

Herberto Helder (1930-) was the co-organizer, with Aragão, of the first review of experimental poetry in Portugal, the Cadernos da Poesia Experimental (1964). His initial experimental works are mainly influenced by the combinatory technique. In the postscript of Electronicolírica, ¹⁰ he mentions an experience that Italian poet Nanni Balestrini did in 1961, with early computer-generated poetry (“Tape Mark I”). Balestrini, Helder reminds us, “choosing some fragments of ancient and modern texts, has supplied them to an electronic calculator that, with them, organized, according to certain combinatory rules previously established, 3002 combinations, later selected”. Helder could have mentioned the OULIPO experiences as well. In any case, he did not have access to a computer, like Balestrini did, neither did he have a literary context in which to develop his ideas. And so, he practiced this technique in different ways. In his poem “Máquina de emaranhar paisagens”, Helder freely combines fragments of the books of Genesis and Apocalypse, as well as François Villon’s, Dante’s, Camões’ and his own, by using combinatory processes, inter and intra-textual, of a great complexity. And he clarifies that his method of poetic creation is stimulated by a process of transference and combination, one from which springs a similarity with “some magic primitive texts, some popular poetry, some medieval lyricism”,¹¹ thus creating a peculiar “ritual magic formula, of which the popular refrain and the medieval parallelism is a trace, as the ballads of the poetry of the Cancioneiros seem to illustrate”.¹² Helder ends up concluding that “(the) combinatory principle is, indeed, the linguistic base of all poetic creation”.¹³

These works of Herberto Hélder were expanded by Pedro Barbosa (1950-) in some of his experiments with computer-generated poetry. Barbosa has been developing poetry with the help of computers since 1976.¹⁴ A pioneer of computer-generated literature, he has had, for the realization of his virtual texts, the collaboration of several programmers, but in all of his projects the literary algorithms are of his authorship. In harmony with the movement of dialogic relations which we have referred, among his first cybernetic poems one can find links to the experiments abovementioned, namely the development of the “Máquina de Emaranhar Paisagens”. Pedro Barbosa was proposing a “renovation of the literary experimentalism in computer-generated literature”, envisaging the processes of transformation and transposition that the transition of the printed to the digital entails.¹⁵

At the same time, his resulting virtual text is an open work: a work in movement, incomplete and indeterminate. As with all generative literature, the virtual represents the possibility to have many different results from the same code. And there is also a place for the reader to interact with the text-program, by means of an active readership.

Humus by Herberto Helder

The first publication of Herberto Helder’s Humus has the suggestive subtitle of "Montage-poem", and in truth this is an intertextual work which uses and transforms the “words, sentences, fragments, images, [and] metaphors” of the homonymous text of novelist Raul Brandão. Helder explains this type of re-writing as archetypal: “there is only one law that includes so much the world of the things as that of the imagination. That law is the law of metamorphosis.” (1964: 5)

This intertextual poem is currently being recreated as a set of virtual texts programmed in Actionscript which are dynamically animated, by means of random substitutions, using the original lexicon of Raul Brandão’s Húmus, a text which was studied and classified in terms of statistical frequency and grammatical position. These lists, which constitute a digital fenotext, are indexed in xml lists of words, which are accessible to the reader in many different ways.

However, this is not merely a combinatory recreation of the literary algorithm conceived (or imagined) by Helder, given that the permutation of the text is also inscribed in the sound generation and in the visual matrix. Every time that a word changes on the screen, a sound and an image are also transformed into their algorithmic equivalent new self, thus becoming a triple layer process: a calculator of interconnected possibilities of text, sound and image. This medial self-reference is an appeal to the concrete materiality of language, and the verbivocovisual expressiveness of language is explored by hypermedial means. But, perhaps most importantly, this recreation aims at being an immersive experience, exploring intermediality, through its extensive usage of integrated multimedia data within interface programming.

The reader can recreate these lists of words, adapting the poem to his/her own taste; s/he can also send, whenever desired, the achieved realisations to a php server, where they are recorded. This process of interactivity is based on networking, granting the recreation with a possibility of collaborative writing and consequently of collective knowledge. Moreover, as the several variations and versions of virtual readers are archived on web servers, it is possible to send this data to rss feeders, PDAs, cellular phones, instant messengers, or through podCasting.

 

Oskarine by Ulrike Gabriel & Oskar Pastior

 

In 2003 software artist Ulrike Gabriel (1964-) created the Oskarine, a project which takes in the poetry of the Romanian-German writer Oskar Pastior (1927-2006). Oskar Pastior counts as one of the most fascinating poets of German language of the 20^th century. The dominant principle of his writing is the restricting rule of a poetic algorithm which Pastior conceived to be textual generator. Accordingly, he renewed various old and partly forgotten forms of combinatory poetics such as the anagram, the palindrome, the sonnet, the sestine, the pantum or the villanelle. It isn’t a surprise that Pastior was the only German speaking member of the international OULIPO group. As a complement of the scriptural aspect of his work, there is also sound: Pastior’s poems develop in a quite specific way by oral presentation, and Pastior’s voice, with its Romanian-German accent and its individual rhythm and melody, has distinctively coined the reception of his oeuvre.

All these aspects play an important role in the Oskarine project which Ulrike Gabriel developed in close collaboration with Oskar Pastior. Oskarine is a poetry machine which generates virtual Pastior poems and displays them with his voice. Gabriel selected several of his poems with strict algorithms. These are especially the forms of the pantum, the sestine, and the sonnet which were programmed in Java. The pantum, for example, is a poetic form that came from Malaysia to France about 150 years ago. It has an open amount of stanzas of four lines each. The second and the fourth line of a stanza appear as the first and the third line of the following stanza, while the last stanza of the poem includes line 1 and 3 of the first stanza as line 2 and 4 to complete the permutation. Pastior recited his particular poem which was recorded and then cut into the units for permutation, that is to say, into audio samples. Following the respective algorithm, the program creates numerous variations of the poem, as well as its reading, respectively. At the same time, the program displays each selected text unit on the screen as a nearly transparent animated line being joined by the following lines. As the permutation continues, this visual text becomes denser and more opaque, thus transforming itself into a scriptural image.

The interest of Gabriel and other software artists in Pastior’s poems is obviously caused by an impression of affinity. This concerns both the self-reference and the processuality of the self-executing text. Another crucial point is the old question of how close the machine could come to the poet, and in exactly which respect we have to realise the differences. The simulative aspect of the Oskarine appears to be remarkable in the first moment. However, we have to state important differences concerning both the methods and the concepts.

We observe an autopoetic mechanism inscribed into the pantum, the sestine, or the sonnet. Pastior is interested in exposing himself as a subject (in the ambivalent sense of being self and submission) to this auto-activity of language and text. Under such constraint, his writing process aims at finding formulations which are as fine and ingenious as possible. This is part of the play between determinacy and indeterminacy. As distinct to the machine concept, it is not decisive for Pastior to program the algorithm correctly. By using this algorithm, he rather tries to find and combine letters, sounds, and words – often neologisms or new elements of the individual Pastior’s language – in such a way that they fit together tonally, visually, and semantically. This process results in the creation of surprising relations and tensions. For example, while the Oskarine has to polish the cuts between the samples as clean as possible, Pastior concentrates on such cuts in order to put them into tension and reflection. While the Oskarine is constantly and strictly or stubbornly permuting the units which were been pre-fabricated by Pastior, such a process certainly belongs to the creating process of the author, but for him it is just a method to achieve a consistent and unique result. He rather uses the algorithm with the intention of inventing various different poetic forms, and by working with additional rules within a cycle of poems. In brief, the machine has a lack of imagination and it is missing the unique modulation of an individual reading.

However, to offer such a reflection on differences proves the impressive aesthetic value of the Oskarine. Conceptually, the project doesn’t just want to simulate the author, but to stage a game between the similarity and difference of two different generative processes, between program and poem, between machine and author as precise as possible. The visual poetry of the Oskarine is symbolising this game like a commentary, by intermedially changing between word and image, chaos and order. Compared to Pastior, Gabriel’s artistic concept is situated on another, more abstract level of observation, but it has to do with similar questions. In the hypermedial composition of various codes, it also presents a great mastery. In our view, this conceptual, as well as formal, density proves this project to be a first-rank artwork on its own. And beyond, it is an exceptional reading and interpretation of Pastior’s poetry.

Stochastische Texte by Theo Lutz and Free Lutz by Johannes Auer

 

The very first computer generated text was presented in 1959. Theo Lutz (1932-), a young student of informatics at the University of Technology (TH) Stuttgart, created what he called Stochastische Texte (Stochastic Texts) using the Zuse Z 22 electronic mainframe. Lutz published the result of his project as an essay in augenblick, ¹⁶ a journal of aesthetics edited by his professor Max Bense (1910-1990). Bense was a professor of philosophy and science theory as well as a writer. Since 1950 he published pioneering contributions to a rationalistic theory of art based on semiotics and information theory.¹⁷ His text aesthetics from 1960 present the conceptual context in which the Stochastic Texts were developed: the turn from idealistic subjectivity to rationalism and objectivity of art, to a programming of the beautiful, the turn from mystic creation to statistic innovation by “a realisation whose existence and perception is based on throwing dice, on a selection of frequencies. One does not change from nothing into being but from the disorder of high entropy into the order of high information.”¹⁸

These poetics were concretely realised by Theo Lutz. He programmed the “throw of dice” as a random generator for the Z 22 which produced grammatically correct “elementary sentences” from a storage of linguistic variables. The machine stores a certain number of subjects, predicates, logical operators, logical constants and the word “IST” (engl.: “IS”), coded as binary numbers. In several steps the program calculates various random numbers which determine the selection of stored variables to generate each sentence. The sentences were first coded as ticker tape and then printed on a telex machine.

The nouns and adjectives stored in the machine were taken from Franz Kafka’s famous novel Das Schloss (The Castle). It is quite interesting that neither this pre-decision was reflected conceptually nor the fact that Lutz had corrected some miss-spellings of the telex printout for the publication in augenblick. We suggest to call this blind spot a latently subjective and romantic factor.

Here we see the starting point of the Free Lutz project (2005) in which net-artist Johannes Auer (1967-) transformed the Stochastic Texts. First of all Auer exactly re-programmed the generator in php. The processing of elementary sentences as they were generated on the Z 22 can now be followed on the Internet.¹⁹ However, for Free Lutz²⁰ Auer conceived a series of extensions. The project allows the generation of sentences with or without the mistakes that Lutz had corrected. Additionally, the audience can interactively invent and write nouns and adjectives and select their gender to be integrated in the program’s writing process. As a result, the mentioned pre-decision which had not been reflected by Bense or Lutz becomes a conceptual part of the Free Lutz project. Hence two different forms of random generation are opposed and networked: the calculating program and the incalculable writing activity of the participating internet users. This aspect is supplied with another feature: Search entries which Internet users enter into the search engine Fireball are fed into the vocabulary of the generator in real time. Thus another text generator – the search engine including its users and the Internet as an open writing process – is self-referentially integrated into the project. As follows, the project presents itself as complex networking of different forms of text generation. Finally, Auer conceived another interesting extension when he invited an actor to recite the sentences as soon as they appear on the screen during a performance or a live radio broadcasting. In this way the processuality is extended to an interaction between actor and audience, especially if the users would enter words with a special phonetic or semantic aspect like, for example, words with a difficult pronunciation or an obscene meaning.

By his project Auer not only honoured and made accessible the Stochastic Texts, but he also offers to reflect the first computer text in relation to present Internet culture. The project conceives the contemporary complexity of digital culture by the recursive networking of several different forms of text generation and presentation. Aesthetically we realise a conceptual fusion of algorithmic and net poetics. It is interesting that Auer consciously integrates the subjective factor again. In doing so he raises a similar reflection like Ulrike Gabriel’s Oskarine. In Free Lutz this reflexive subjectivity is realised when text is generated by the performance audience or the Internet users or, increased, as say “authentification” by the physical, vocal presence of the performing actor. This can also be interpreted as a critical, even deconstructive reading of Bense’s aesthetics, as a re-evaluation: The radical calculation and statistical determinacy of the beautiful is opposed by indeterminacy and individuality. This link to Romanticism is dialectically outlined as a latent subtext of Bense’s poetics.

 

Conclusion: Siting the “poetry” in e-poetry

Our discussion on achievements and functions of transformations in(to) the digital results in two perspectives:

Firstly we can observe how principles and methods which have been formulated within the experimental poetics program really urge to be continued, extended, experienced in electronic space. This means that forms and strategies of medial self-reference, processuality, interactivity, intermediality, or networking as outlined by the pre-texts offer to develop their potential under the conditions of electronic writing. This view especially motivates the works of Barbosa and Torres in their intertextual reference to predecessors like Aragão and Helder.

Secondly we have shown that recreations within the program of poetic experiment lead to new interpretations and evaluations. This counts for the poetic conditions of both the pre-text and the project of its transformation. Assuming the general poetic staging of self-reference of language, or of the technological and cultural conditions of language use, we can notice a displacement of abstraction of meta-observations. And we can further notice a displacement of poetic values. In this respect it is interesting that the “subjective factor” (as reflexive subjectivity) becomes part of the game again. And particularly in digital poetics it is interesting to raise the question again: Who is speaking? Who is writing, programming, reading, viewing, reflecting, perceiving, feeling etc. – and under which conditions?

Thus we can observe both: continuation and change simultaneously and complementarily in the poetic domain. From this point of view, we shall providesome final remarks on siting the “poetry” in e-poetry:

Poetry/poiesis is a unique, flexible and dynamic mode of observation including self-observation. It creatively intrudes on its own general autopoiesis, by presentation, reflection, transformation, recreation, etc. Here, poiesis/poetry is the medium in which any creative mechanisms can be simulated and observed. In this sense, poetics also works self-referentially as technology (poietike techné). Particularly where poetics adapts itself to the culture of technology and relates this to its own means and procedures. It experiments with experiences on how technology literally makes sense between determinacy and indeterminacy, restriction and openness. “Literally” means that poetry, as language art, raises the linguisticality of the technology culture, as well as existing and new language techniques, and acts these out, condenses and recontextualizes them—in short, contaminates them poetically.

Poetry no longer progresses, instead, it constantly extends its possibilities. Attributes like e-poetry, digital, cyber, visual, sound poetry etc. represent this. These are open concepts of orientation with which art communications organise themselves in factual, social and temporal ways. However, genres, or object areas that are clearly, aesthetically outlined, can hardly be defined by these concepts. Even if their academic use occasionally tries at standardisation and institutionalisation.

Poetry, as language art, is—in keeping with Umberto Eco—art in movement, inspired by a (renewed) performative turn. With this performance, borders within art, as well as other communication and sense systems are constantly moved, and the works of art and their protagonists are process orientated. In this sense poetry is not “progressive” but rather “prozessive Universalpoesie” (‘processive universal poetry’).