|
Recombinant text
|
|
Recombinant text is a term proposed by Michael Allan to describe a collaborative medium of design and composition based on a mechanism of evolutionary genetics. In theory, recombinant text works much like a system of natural populations, or evolutionary algorithms; but instead of running automatically, the mechanism is human controlled, and directed toward the design of cultural artifacts. Implementations have been developed for the purpose of collaborative writing. In typical use, a text is replicated into a population of variants, one per author; authors swap fragments and arrangements of the text, from peer to peer; and genes encode a traceable record of authorship throughout. Communication pattern Compared to other collaborative media, recombinant text has a distinctive communication pattern:
Unlike the centralized pattern of most other collaborative media (such as Wikis), recombinant text has a distributed, peer-to-peer pattern. Instead of a central copy of the shared design or composition, the text has multiple copies, one per author. Instead of pushing to the center, authors pull contributions from peer to peer. As a result, the text diverges into multiple variations that co-exist side by side. Collectively, the text has spatial diversity. In statistical terms, it has a population. Typical characteristics The text exists as a population of variants, one per author. This is a natural consequence of the distributed, peer-to-peer communication pattern (above). The resulting population forms as a collection of documents posted at various locations on the Web. All documents are related, either by common origin or by cross-transfer. Each is the working copy of a single author, who has independent creative and editorial freedom. Authors swap fragments and arrangements of text, peer to peer. This is how authors collaborate, by pulling contributions from each other. Different methods of transfer are possible. The most basic is like copy and paste - an author selects part of another author's work, and transfers it to his own. But recombinant transfer has advantages over ordinary copy and paste, because it is aided by genetic alignment, a feature that reveals patterns of similarity among texts, and helps to guide each transfer. As well, the source authors are automatically acknowledged in the target text, with each transfer. A genetic code is a set of rules for encoding and decoding recombinant texts. Current implementations use a form of XHTML, extended with genetic structures. The most important structures are genes. A gene is a datum encoding a part of the text. Genes serve as: * atoms of communication, to convey compositional ideas from author to author; * building blocks, to physically compose and re-compose the text; and * conservative elements, to maintain the fidelity of information and its authorship in a changing population. A population is a community of individual members that exchange genetic information. The individuals are separate copies of the text. Each copy is a "working document", associated with a single author or editor. Mutation is essentially text editing. Each author has independent editorial freedom over his own working document. Consequently the population acquires variant documents, and the gene pool acquires variant genes. These variations accommodate diversity. They are, in turn, grist for the mill of recombination. Recombination allows an author to improve a working document by transferring select bits of text from the documents of other authors. On the surface, recombination can be as simple as copy and paste; underneath, it involves the transfer of discrete genes. Use cases Several approaches to recombinant text have been explored to date, each corresponding to a different use case, or mode of use. Paired-regions This is where you are reading another author's document, and discover a desirable piece in it. If you like, you can copy it, and paste it into your own working document; either as a new insert, or as replacement for prior text. And then continue reading. Current development of textbender is focused on this approach. Evolutionary genetic systems The theoretical underpinning of recombinant text is evolutionary genetics - the synthesis of Darwinian evolution and modern genetics. Aside from its significance as a biological theory, evolutionary genetics also has practical applications in fields such agriculture, engineering and art, where it underpins systems that combine optimization and search in detail, with creative composition and diverse expression in the large. Recombinant text is one such system, and the following table places it in context. These differences detract from the comparison. Oral Tradition There is a notion too, that recombinant text may be a re-creation, in a sense, of rhapsodic verse in the oral tradition. Substitute the genes of a recombinant text for the formulae, epithets and stock phrases that were the raw materials of improvised recital - what Ong calls its "prefabricated parts"; and substitute the Web for the wandering singers who formed the collective memory and communication medium of the oral tradition; and it appears that recombinant texts are composed and re-composed from posting to posting, much as songs were once "stitched together" from recital to recital. In this sense, a recombinant text re-creates an environment similar to the one in which the ancient epics evolved; a fluid dynamic and dialogue among artists, long since frozen by the invention of writing. This proposal has led, in turn, to the design of an open electoral system for the general purpose of advancing candidates for public office, policies for executive action, and legislative bills for statutory law. Its development has since commenced as "project Votorola". 
|
| < Prev | Next > |
|---|
Main Menu
| Home |
| Articles |
| Search |
| Submit Article |
| Contact Us |
| Donate to Wikibin |
