MaMoPo : by : PoLiOu
Machine Modulated Poetry by Potential Literary Outlaws [NB]
This piece was written for the 'Writing and Computers Newsletter',
details from the editor, Mike Sharples at : firstname.lastname@example.org.
Hypertext, as literature, is still in its
infancy. The majority of 'serious' [NB] literary hypertexts which have thus far been produced are fiction, and
are composed chiefly in the ever-more familiar link-node structures which proliferate
on the World Wide Web. [NB] In this form, pieces of prose (rarely poetry) of variable length, now generally
referred to as 'lexia', are stored in 'spaces' which usually correspond to text windows
on the screen, as well as to the 'nodes' of the link-node model. The author uses
her chosen software to add 'links' from one node to another. Text spaces may be linked
to one or more other lexia. Words, phrases or sentences (any smaller linguistic structures)
within each lexia are also commonly linked. The details of what is possible depend
on the software used, and systems vary a great deal. Some are very flexible and sophisticated.
We are beginning to see, for instance, 'multi-head multi-tail' linking.
[NB] The end result
is an emergent medium, more familiar in the world of technical writing and expert
systems, but with great literary potential; a medium which gives both author and
reader new freedoms, resources and responsibilities.
Poetry and technology is not a customary rhyme. As compared with
fiction writers, there are even fewer practising poets making use of the new potential
offered by hypertext and related systems for the composition and publication of poetic
literature. Poets have been relatively slow to familiarize themselves with the new
technology other than through the now near-universal use of the word processor. For
most poets 'computer-plus-software' is not a new medium but just a more convenient
and flexible simulacrum, which refers to the traditional tools of pen and paper.
But apart from a (romantic or levelling) suspicion of new technology,
there may be other reasons for the poets' relative disengagement. Some of the new
literary hypertexts use a language (within their lexia) which could be characterized
as 'poetic', however the medium does not in itself encourage or enforce that
more profound involvement with the nuts and bolts of language which is the poet's
stock in trade. A poem may be composed within a 'space' in a larger hypertext without
any effect on the processes of its composition or its reading. In this case, the
hypertextual system has not directly engaged or influenced its poetics, and the lexia/poem
might just as well have been found on the leaf or leaves of a paper codex.
On the other hand, the parts of a poem could be dispersed across
a number of spaces in a link-node structure. Here, the system has a real role to
play. Authors' decisions as to how to link and where to place the parts, and readers'
decisions about what to read and when, have the potential to generate new varieties
of poetic experience. If the software used has the facility to automate or animate
certain readings of the structure, then the poet is able to compose, for example,
multiple real-time (silent) readings in a form possessing resources which are radically
different from those of the book. At this point the temptation to introduce ornamentation
from other media - pictures, audio, video - becomes all but irresistible. Most writers,
especially new writers, will succumb.
My own first explorations of machine modulated poetry were in the
field of animated or kinetic presentations of previously composed
work, or of work designed to have multiple reading paths, [NB] although even at this stage (late 1970s),
my basic unit was the word, not the lexia or textual space of link-node hypertext.
Many poets will be attracted to the possibility of presenting their work kinetically.
Scoring a previously composed text (setting the pauses between and speed of display
for words, phrases and lines) according to how the poet inwardly 'hears' the composition
(without having to trust to the vagaries of punctuation or avant-garde lineation)
has obvious appeal. One of the better-known practitioners in this field in the US
is Robert Kendall, who is also running an Internet correspondence
course on 'Hypertext Poetry' for the New School (NY). [NB]
However, scoring previously composed material for presentation
on the screen is a rather limited or constrained use of the programmable machine.
To myself and others, the possibilities for active reading along multiple or indeterminate
pathways holds greater attractions, opening up to literature a potential for real-time
process art. It is not as if such piratical literary experiments have not been attempted
before. William Burroughs and Brion Gysin produced their 'cut-up' works in post-war
Paris; Jackson Mac Low in the United States has created a substantial and important
body of process-based work, as have certain other artists of the Fluxus group, particularly
Emmett Williams; the 'OuLiPo' (Ouvroir de Litterature Potentielle, founded in 1961
by Raymond Queneau; best-known members - to English readers - Georges Perec and Harry
Mathews) continues to experiment with 'constrictive' and generative form; the well-known
critic, Hugh Kenner and Joseph O'Rourke produced an interesting program called 'Travesty'
(Byte , Nov. 1984) which modulated any piece of language fed into in it by
reproducing patterns of letters or words which the program found in a text at different
user-definable levels or 'orders'; and more recently, C. O.
Hartman, also a poet and critic, programmed varieties of Jackson Mac Low's 'diastic'
procedures which, in concert with Travesty, have been used to produce his Sentences
(Los Angeles: Sun & Moon, 1995). [NB] More generally, innovative and radical contemporary poetics, especially
those associated with and promoted by the (US-centred, so-called) 'Language' cohort,
have encouraged an explicit interest, not only in the rhetorical, but also the hard,
grammatical elements of new poetries.
It is in this context that my own work has developed, often in
ignorance of parallel or previous experiments elsewhere. My work - a project I have
given the general title 'Indra's Net or Hologography' - is distinct from more familiar
hypertext in a number of respects. The underlying texts (which I call my 'given'
texts) are engaged with the contemporary poetries I have just cited. I see my writing
(in so far as it is composed prior to any generative process and in so far
as it aims to produce a generated text of a certain kind) as a part of this particular
contemporary poetic practice. My 'nodes' are not the usual 'size' of lexia. They
are most frequently word sized. My 'links' are self-generating, rule-based and determined
by relatively simple linguistic structures. This means both that the generated pieces
disrupt and interrogate language itself in a way that hypertext usually does not,
and also that the 'link-node' model loses uses its usefulness as a description for
this type of work. Indra's Net pieces employ generative algorithms and semi-aleatory
processes and the composition of the algorithm is seen as an integral if normally
invisible part of the composition of the piece. One of the unique facilities offered
by the computer in this context is the ability to set up a feed-back loop. 'Experimental'
texts can be generated and the results reviewed quickly and painlessly enough to
allow the processes to be modified and improved. Once distributed, the pieces 'run'
and generate text for a reader. The reader can interact but does not choose pathways
between words directly in the way that she might choose a pathway through the spaces
of hypertext fiction. However in my most recent distributed piece, readers can alter
the work itself (irreversibly), collecting generated lines or phrases for themselves
and adding them to the hidden given text so that eventually their selections come
to dominate the generative process. The reader's copy may then
reach a state of chaotic stability, strangely attracted to one particular modulated
reading of its original seed text. [NB] Work in progress is towards a series of '(Plastic) Literary Objects' which
will be both generative and responsive, triggered by as many as possible of the program-
and user-generated events which are accessible using a standard computer system.
This latest object will be a far cry from the average web page.
One of the principles guiding my work is a belief in the distinctive
qualities and special value of a literary experience characterized by silent reading.
In the new culture of cyberspace this experience must strive for self-preservation
in the face of a multimedia assault which seems sometimes to claim that literature
is next to nothing without its loud and colourful ornamentation, or that literature
is doomed to become an ornament for its post-literal content. It is true that literature
will only thrive on 'the matrix' if it can develop in ways which fully exploit the
capabilities of a new cybernetic infrastructure. Literary hypertext generally, and
the more radical machine modulated poetries have an important role to play in these