Software Art and Writing

> Software Art and Writing
> Florian Cramer, drafted with Ulrike Gabriel
> 5/18/2001
> What is software art? How can ``software'' be generally defined? We
> had to nsweer these questions at least provisionally when we were
> asked to be with the artist-programmer John Simon jr. in the jury of
> the ''artistic software`` award for the transmediale.01 art festival
> in Berlin, Germany.
> Since more than a decade, festivals, awards, exhibitions and
> publications exist for various forms of computer art: computer music,
> computer graphics, electronic literature, Net Art and
> computer-controlled interactive installations, to name only a few,
> each of them with its own institutions and discourse. Classifications
> like the above show that attention is usually being paid to how, i.e.
> in which medium, digital artworks present themselves to the audience,
> externally.

One assumes that if the above mentioned catergorization actually meant something worthwhile, then the basic paradigms of software art would be understood -and Cramer's article would not be neccessary.

In fact, we are confrontated by a plethora of undefined terms which everybody uses in ways which gives themselves the best political advantage in the ideological battle for hearts and minds -and above all funds.

Art historians develop their careers by playing with these terms. However, in the meantime art practice suffers because either artists accept the theoretical straight-jacket offered by the (somewhat confused) theorists -or are excluded from the system because the theorists are increasingly determining the debate -not only in theoretical terms but also in practical terms -as they continue increase their power in the artistic education, funding and presentation system.

> They also show that digital art is traditionally
> considered to be a part of ``[new] media art,'' a term which covers
> analog and digital media alike and is historically rooted in video
> art. But isn't it a false assumption that digital art - i.e. art that
> is consists of zeros and ones - was derived from video art, only
> because computer data is conventionally visualized on screens?

Indeed it is not only a false assumption -it is also a historical falsfication to assume that digital art developed out of video art!

In fact, the history of "computer art" (i.e. the use of computers by artists to produce art) goes back to the 1950's and 60's -before video was on the market and therefore before video art could start. One wonders how such an important historical error has pesisted so long in contemporary art theory.

One also wonders how long "new media" needs to be around before somebody discovers a more intelligent term -and produces a more intelligent analysis of how "electronic art" might relate to non-electronic art (concidering both simularities and differencea), how "digital art" might differ from non-digital art -and especially -seriously concider the consequences of artists programming their own work as opposed to being consumers of commercial products.

> By calling digital art ``[new] media art,'' public perception has
> focused the zeros and ones as formatted into particular visual,
> acoustic and tactile media, rather than structures of programming.
> This view is reinforced by the fact that the algorithms employed to
> generate and manipulate computer music, computer graphics, digital
> text are frequently if not in most cases invisible to the audience.

Perhaps this is not particularly revolutionary (or even unusual). Traditional music, ballet, film or even literature does not often expose its internal algorythims directly to the public either. Interested parties are often expected to deduce the underlying algorithm through their knowledge of the medium, the work, the artist and the artist's personal history. Presumably, a large part of the work of understanding the underlying algorithm is dependant on the knowledge the viewer.

What is perhaps really worrying is the fact that that the algorithms employed to produce so called "digital art" are often unknown and/or ununderstood by the artists themselves -let alone the viewer!

Interestingly (and perhaps significantly) some modern theories of literature apparently reject completely the complex nexus of influences which act upon the the creation of the work -apparently depriving literature of all meaning except that which the individual observer/reader wishes to impart to it. This attitude also seems to be adopted by many in the visual arts community -it is also paralleled by a theory which claims that the physical medium of a work is unimportant and that only the concept is important. One would have thought that these two theories were incompatable (one theory denying the value of the artist's intention and the other theory promoting the artist's intention to a godlike status). However, in practice, it seems that art theory conciders these two theories as being supportive of each other.

> While the history of computer art still is short,

What is short? It is perhaps no older than a single human lifespan -but nevertheless, there are young artists active today who were born after the birth of computer art -and indeed most of the academic community (armed with phd's) who are responsible for educating the next generation (or for funding and organising the dissemination of contemporary art) have not personally experienced the history of computer art since its primitive beginnings.

Computer art theory is apparently based on theoretical reconstructions and is not developed on the basis of historical experience. In practice, many of the original pioneers seem to have been marginalised by contemporary usurpers. This practice seems to go further than the usual dialogue between successive generations.

> it is rich with
> works whose programming resides in black boxes or is considered to be
> just a preparatory behind-the-scenes process for a finished (and
> finite) work on CD, in a book or in the Internet.

This is perhaps true -but then it might be worthwhile enquiring into the historical reasons for this. In fact, in my personal experience, in the 1980's commercial systems were conciously promoted by the new propagandists -who seem to ignore (as much as possible) the early pioneers -despite the fact that it was the efforts of obscure pioneers (not only in computer use -but also in music and visual art in general) who provided the inspiration for much of the software which was sold to an ignorant public.

> This is case with
> practically all books and newspapers in their dependence on word
> processing and typesetting software, and it is the case all audio CDs,
> even (and particularly) those which contain algorithmically generated
> music.

Perhaps this is true too -but again, it may be worthwhile asking how this came about. The realationship between word processing and typesetting and visual art is not particularly obvious -but even within the world of academic publishing it is worthwhile to note that the original de facto standard for academic publication was UNIX -which later became cloned into Linux -and yet many (if not most) academic institutions and publishers demand the use of commercial products -despite the indictment of the producer of these products for dubious commercial practices.

> The distribution of John Cage's sound play ``Roarotorio,'' for
> example, includes a book, a CD and excerpts of the score, but not even
> a fragment of the computer program which was employed to compute the
> score.

Well, until now -most of contemporary digital art theory has done its best to promote the idea that "what goes on under the motor cap" is totally irrelevant. So I suppose it is an interesting intellectual game to start promoting the opposite.

> While software, i.e. algorithmic programming code, is inevitably at
> work in all art that is digitally produced and reproduced, it has a
> long history of being overlooked as a conceptual and aesthetic factor.

Not only overlooked -it has been deliberately downplayed (prossibly to support the commercial and personal intersts of those involved).

Even so -we may also be suspicious of equating "software" with "programming code".

> This history is paralleled in the evolution of computing from systems
> that could only be used by programmers to systems like the Macintosh
> and Windows which, by their graphical user interface, camouflaged the
> mere fact that they are running on program code, in their operation as
> well as in their aesthetics. Despite this history, we were surprised
> that the 2001 transmediale award for software art was not only the
> first of its kind at this particular art festival, but as it seems the
> first of its kind at all.

How on earth can this be true?

Almost by definition -the early pioneers of computer art were experimenting before the availability of commercial software for the production and manipualtion of sound and image. By definition -all computer art pioneers were what one might now call (for want of a better definition) "software artists". Therefore, all production, presentations, festivals and awards concerning computer art (before the 1980's) were concerned with 'software art". I seem to remember that Siggraph goes back to the 60's -so the belief that "software art" is something new must clearly be a historical error of the greatest magnitide!

> When the London-based digital arts project I/O/D released an
> experimental World Wide Web browser, the Web Stalker
>, in 1997, the work was perceived to be a
> piece of Net Art.

Yes -and .net art has consistently promoted the importance of content as opposed to algorythmic structure.

> Instead of rendering Web sites as smoothly formatted
> pages, the Web Stalker displayed their internal control codes and
> visualized their link structure.

So is the author now claiming that software/computer art derives from Web Stalker?

> By making the Web unreadable in
> conventional terms, the program made it readable in its underlying
> code. It made its users aware that digital signs are structural
> hybrids of internal code and an external display that arbitrarily
> depends on algorithmic formatting. What's more, these displays are
> generated by other code: The Code of the Web Stalker may dismantle the
> code of the Web, but does so by formatting it into just another
> display, a display which just pretends to ``be'' the code itself. The
> Web Stalker can be read as a piece of Net Art which critically
> examines its medium. But it's also a reflection of how reality is
> shaped by software, by the way code processes code.

Perhaps this is correct for web pages -but maybe the asumptions made here also demonstrate of  total misunderstanding of the relationship between constructing the html code for a web page and the nature of programming -and language in general.

The esential characteristic of "code" is that it is a reversable process. One codes and one decodes and any difference between the coded input and the decoded output reflect an error (or at least an infidelity) in coding.

Speech can be coded by writing -an image can be coded into jpeg or some other format -and a graphic layout can be (almost) coded into html. One says "almost" because in fact there has been a conflict between the "formatting" aspect of html -which allows "information" *whatever that may be) to be "represented" on any computer using html to "re-present" the data. The original html specifications were therefore rather summary -and offered only minimal formatting requirements -so that in practice the person creating the html document had little control over its actual presentation on the specific system used by the person viewing the page.

Historically -the inexactitude implicit in the way the html formatting system was interpeted by the individual's computer system was unacceptable to artists and graphic designers who wanted complete control over the presentaion of their product (although the lack of control might also have offered an interesting challenge to conventional wisdom concerning the importance of the need to present these products as close to their creators original intentions as was possible.

As a result of the apparent need to control the final presentation -html shifted from being a "formatting" system (concerned with the structuring of data for presentation purposes) to a coding system (concerned with the correct reproduction of an original).

So assumimg that we can reconstruct speech from writing -and a graphic layout from html -then it is legitimate to refer to writing and html as "code".

However, one can easily question the belief in "language" as a "code": A language has a repertoir of basic symbols, a grammar which controls the system for generating compound symbols (or strings) -and possibly, a system for deriving some significance from the grammatically correct constructions of the grammar. The realtionship between the construction of grammatical correct compounds and the (correct or incorrect) interpretation of these compounds is still subject to speculation -but one thing is clear -there does not seem to be a direct one-to-one relationship between "intepretation" and the "grammatically correct construct".

Wittgenstein has suggested that meaning is not external to language -but an inherent aspect of its internal structure -implying that one cannot communicate via (a) language anything which is not part of (that) language (or derivable from, or within it). As far as I know -this has never been formally repudiated -and if it is correct -we may assume that there is no "thought" which is external to language and which can be "coded" by language. In any case, we can observe that it is impossible to compare most expressions of language with the original that produced them. By definition, language can never be "code". Even (apparently non-ambiguous) computer programmes are difficult to compare with the original brain patterns of the person who created them.

One may therefore assume that the 20th century linking of "language" to the concept of "code" was the largest intellectual error of the century -equivalent to the 19th centurary belief in phrenology -or the ether.

> Since software is machine control code, it follows that digital media
> are, literally, written. Electronic literature therefore is not simply
> text, or hybrids of text and other media, circulating in computer
> networks. If ``literature'' can be defined as something that is made
> up by letters, the program code, software protocols and file formats
> of computer networks constitute a literature whose underlying alphabet
> is zeros and ones.

That must be one of the most stupid remarks ever written by anyone claiming to have any understanding of either computers or literature!

If '``literature'' can be defined as something that is made
 up by letters' -then indeed a monkey with a typewriter can produce brilliant literature -however, I suspect most people would assume that literature operates on a higher level of analysis than one based on letters! More sophisticated thinkers might also understand that the representation of language by symbols is not identical to language itself. One is a static representation -and the other is a dynamic process of creation. One should not confuse the moon with the finger that points to the moon!

> By running code on itself, this code gets
> constantly transformed into higher-level, human-readable alphabets of
> alphanumeric letters, graphic pixels and other signifiers.

I suspect that this "code" originated as a "higher-level, human-readable" language -and not simply on the aphabetical level. Most adults operate at least with words, if not sentences -and sometimes entire paragraphs -and do not operate on the level of individual letters.

If literature (or indeed even computer art) operated on the level of letters -then presumably simply re-coding the alphabet would be sufficient to "translate" a Russian book into English. One understands that in practice this is not so.

> These
> signifiers flow forth and back from one aggregation and format to
> another.

Which "signifiers"? The author hardly seems to understand the diffences between a "format", a "code" and a "language" -apparently completely ignores the importance of grammar and is now apparently basing their argument on yet another undefined (and probably misunderstood) concept.

> Computer programs are a literature in a highly elaborate
> syntax of multiple, mutually interdependent layers of code.

If we substitute "language" for "code" then perhaps this is the most (and almost only) correct statement made so far in the text. However, until now, the author seems to be completely unaware of the implications of the difference between code and language.

> This
> literature does not only rely on computer systems as transport media,
> but actively manipulates them when it is machine instructions. The
> difference is obvious when comparing a conventional E-Mail message
> with an E-Mail virus: Although both are short pieces of textwhose
> alphabets are the same, the virus contains machine control syntax,
> code that interferes with the (coded) system it gets sent to. It could
> be compared to the poisoned pages of Aristotles ``Poetics'' in Umberto
> Eco's novel ``The Name of the Rose,'' with the difference that in
> computer viruses, the mere language induces the lethal dose.

Well, I supose these remarks once again reflect the confusion between "language" (the grammatical construction of potentially significant compound constructs), "literature" (written language -possibly of an artistic nature) and "code" (the reversable substitution of representational systems).

In the example above -it is true that "literature" does not usually have a dynamic influence on the book -while "computer code" may well influence the computer which interprets it. However, this is an oversimplification which confuses many different things. Traditional literature no longer uses hand set type to produce the book -so the codes used by the typesetting machine could be conciderd similar to the codes used to represent a (static) html page.

Admittedley, the situation (comparason) gets more complex when one starts concidering the use of dynamic programming languages (such as Java) via the net. In this case, the programme may influence the machine in a more fundamental way than is possible by html. However, in practice, the effect of the programme on the machine is limited for security reasons (well, in theory it is restricted -the practice it is slightly different -as any Microsoft user knows when subject to attack by the latest virus).

So -under some circumstances -the binary code can (if the computer is running a programme that responds to the code) affect (in intended or unintended ways) the behaviour of the computer upon which it is running. However, this is somwhat of a tautology -because the generally accepted definition of a computer is a machine which is controlled by a programme stored in its own internal memory.

On the other hand, we might also conclude that (at least according to western philosophical tradition) human beings also operate on the basis of internally stored programmes -so perhaps humans are computers. If we look at human communication in terms of "language" -and not in terms of the realtionship the letters in literature might have to the book -then we discover that humans use language to modify each others behaviour. If we now concider literature as coded language -and the target of literature/language is another human being and not the book -then we might conclude that the results of ones comparasons are entirely dependant on the level of conceptual analysis upon which the analysis has been made.

Unfortunately, it is exactly the implications of the different levels of conceptual analysis that the author seems to be ignoring -so Cramer's text appears to becomes increasingly meaningless as one progresses through it.

> If programming is writing with machines, software code at once is
> language and structural manipulation of a technical system.

If thought is the (arbitrary) manipulation of symbols -then computers "think" -I suppose by the same logic one could prove that the author of the text on software art also "thinks". Nevertheless -one can seriously question the validity of the result -and therfore the value of "thinking" -either by a machine or by the author.

These are not simply gratuitous insults -they suggest that "thinking" is not an entirely arbitrary manipulation of symbols -but also involves some concideration for the implications of the results of the process -and until now -this understanding seems to be missing in Cramer's text.

> This
> aspect is neither covered by the concept of ``hypertext,'' nor by the
> concept of ``multimedia.'' As umbrella terms for ways of structuring
> and formatting data, they both don't imply by definition that the data
> is digital and that the formatting is algorithmic. Nevertheless, the
> ``Web Stalker'' shows that hypertext and multimedia on the one hand
> and software art on the other are by no means exclusive categories.
> They could be seen as different perspectives, the one focussing
> display, the other one the concept and systemics.

One would have thought that any serious understanding of "software" (or language) would have seen this as being so self-evident that it did not warrent discussion. Only from a position of misunderstanding do these things aquire the significance given to them in this text.

> But is code which technically manipulates systems exclusive to
> computer programming? The history of algorithmic, self-executing
> writing is much older than the history of the computer. Besides
> mathematics proper, it includes the permutational language of the
> Kabbalah, Lullian combinatorics in Renaissance poetry, in Novalis and
> Mallarmé, and combinatory language games of Dada and the French Oulipo
> writers.1. Software code doesn't even have to be algorithmic. If code
> is, mostly simply put, instructions that make up and control a system,
> it is - according to the legal theoretician Lawrence Lessig - law, and
> law vice versa is executable code2. Lessig's equation can be read to
> extend from secular and religious law into the realm of art when we
> consider, for example, the  Composition 1961 No. I, January I by the
> contemporary composer and former Fluxus artist La Monte Young:

So, it now appears (after a long discussion of "software as code") that software is not code -and that software (and presumably softeware art) is primarily concerned with the construction of rule based systems and the implications (intepretation?) of these systems.

Why do we need to be fed a whole load of false theory before attempting to discuss a more intelligent approach?

> This piece can be called a seminal piece of software art because its
> instruction is formal.

If this piece is seminal -and it was produced in 1961 -and it concerns the nature of formal instructions then why did we need to make such a detour to get to it. Apparently, this piece is more representative of "software art" than more contemporary examples -and yet earlier in the text it was implied that software art was a new phenomnon pioneered by the Transmedial festival.

> At the same time, it is extremist in its
> aesthetic consequence, in the implication of infinite space and time
> to be traversed. Unlike in most notational music and written theatre
> plays, its score is not aesthetically detached from its performance.
> The line to be drawn could be even considered a second-layer
> instruction for the act of following it. But as it is practically
> impossible to perform the score physically, it becomes meta-physical,
> conceptual, epistemological.

So is the author now departing from his previous definitions of language as code? Are we now to understand that language is metaphysical, conceptual and epistemological (whatever these words might mean)?)

> As such the piece could serve as a
> paradigm for Henry Flynt's 1961 definition of Concept Art as ``art of
> which the material is `concepts,' as the material of for ex. music is
> sound.''3

Perhaps one might also carefully note the (potential) significance of this remark. It states that "concept" is the material for art -it does not state that art is "immaterial" because it relies on concept!

There is a subtle difference: If concept is "material" -then the "material" of the work remains essential to that work -while if art is "immaterial" then the "matrial" has no significance.

It would seem that not only has modern art theory completely misunderstood the nature of language -it has also (deliberately -or accidentally) completely misunderstood the nature of "concept" and the nature of "material" with respect to artistic practice.

If we then concider the curent (and increasing) importance of art theory for art practice -then we can easilly suspect that current art theory has been extremely damaging to conmtemporary art practice: Powerfully propagating its own erronious theories and ignoring all art practices that might have contradicted these false theories.

> Tracing concept art to artistic formalisms like twelve-tone
> music, Flynt argues that the structure or concept of those artworks
> is, taken for itself, aesthetically more interesting than the product
> of their physical execution. Flynt's Concept Art thus integrates
> mathematics as well, on the acognitive grounds of ``de-emphasiz[ing]''
> its attribution to scientific discovery.4 With this claim, Flynt
> coincides, if oddly, with the most influential contemporary computer
> scientist, Donald E. Knuth. Knuth considers the applied mathematics of
> programming an art and whose famous compendium of algorithms is duely
> titled ``The Art of Computer Programming.''5

Why on earth it should be so odd -I fail to understand. Knuth and Flynt are appartently both within the same tradition -one which has been largely denied and buried as deeply as possible by modern theorists. Only when one does not understand the historical corruption of computer art committed by its apparent protaganists -is there any mystery.

> Should the transmediale software art jury therefore have consisted of
> mathematicians and computer scientists who would have judged the
> entries by the beauty of their code?

Perhaps this depends on how one defines "the beuaty of their code". Is literature judged on the beauty of the authors code -or is judged on the basis of an understanding of the complexity of the language it employs?

Perhaps the software art jury should have consisted of artists who has a thourough understanding of the nature of software art. Unfortunately, such people are difficult to find -and probably have attitudes which undermine the current ideology.
Presuambly, these conceptual differences explain why people with understanding are kept out of view until the current generation of experts can reinvent the wheel.

> What is known as Concept Art today is less rigorous in its
> immaterialism than the art Flynt had in mind. It is noteworthy,
> however, that the first major exhibition of this kind of conceptual
> art was named ``Software'' and confronted art objects actually with
> computer software installations.6. Curated in 1970 by the art critic
> and systems theorist Jack Burnham at the New York Jewish Museum, the
> show was, as Edward A. Shanken suggests, ``predicated on the idea of
> software as a metaphor for art [my emphasis],''7. It therefore
> stressed the cybernetical, social dimension of programmed systems
> rather than, as Flynt, pure structure.

So -if conceptual "software art" was aready being produced in 1961, it predates "conceptual" non-computer art -and yet we are also told that software art is a new phenomenon.

What a bizarre mix-up: "Conceptual" art copies "cybernetic/computer art" and as a result everybody assumes that "digital art/new media" means "immaterial" concept art.....

> Thirty years later, after personal computing became ubiquituous,
> cultural stereotypes of what software is have solidified.

Unfortunately yes -especially as they have apparently solidified around multiple layers of misinterpretation building upon previous layers of misinterpetation..... (or was it misrepresntation building upon layers of misrepresentation?).

> Although the
> expectation that software is, unlike other writing, not an aesthetic,
> but a ``functional tool'' itself is an aesthetic expectation,

The error is apparently not an "aesthetic" expectation but the result of an intense campaign of political propaganda.

> software
> art nevertheless has become less likely to emerge as conceptualist
> clean-room constructs than reacting to these stereotypes.

This would seem to be inevitable -when one conciders the increasing power of artistic dogma over artistic practice.

> The ``Web
> Stalker'' again might be referred to as such a postmodern piece. In a
> similar fashion, the two works picked for the transmediale award,
> Adrian Ward's ``Signwave Auto-Illustrator'' and Netochka Nezvanova's
> ``Nebula M.81,'' are PC user software which acts up against its
> conventional codification, either by mapping internal functions
> against their corresponding signifiers on the user interface
> (Auto-Illustrator) or by mapping the signifiers of program output
> against human readability (Nebula M.81).

Indeed, an intersting example of how "postmodernism" has created false stereotypes and then pretends to argue against them -finally granting itself eternal and universal justification of itself by claiming that all human activity is as fundamnetally pointless as postmodernism's own silly mind games.

> Contrary to the formal language of fixed scores like La Monte Young's,
> but similar to literature in nonformal languages,8 computer software
> can even be programmed to recursively rewrite itself. If software
> coding is writing, it's processual not only as a computation process
> in the machine, but also when it's being composed in a programming
> languages. The works of Ward and Nezvanova and, for example, the
> computer code poems of mez and Alan Sondheim show that coding is a
> highly personal activity. Code can be diaries, poetic, obscure, ironic
> or disruptive, defunct or impossible, it can simulate and disguise, it
> has rhetoric and style, it can be an attitude.

So, first language is code and then language is not code and now language is code again. One wonders whom is trying to fool whom -and why?

> Such attributes might seem to contradict the fact that artistic
> control over combinatory iterations of machine code is limited,
> whether or not the code was self-written. Unlike the Cagean artists of
> the 1960s, the software artists mentioned above seem to appropriate
> this not merely as a means against intention, but as a simultaneous
> negation and extension of the writing subject.9

Unlike the (Cagean?) artists of the 60's (what happened to Flynt?) -contemporary theory is so confused one can only assume that a return to the sixties as starting point may be an essential step towards unraveling the current confusion -assuming of course, that the current high priests of dogma are prepared to face up to their responsibilities as arbitors of contemporary fashion.

> References
> [Bar68]
>      John Barth. Lost in the Funhouse. Anchor Books. Doubleday, New
>      York, London, Toronto, Sydney, Auckland, 1988 (1968).
> [Cra99]
>      Florian Cramer. Permutationen, 1996-99.
>      cantsin/index.cgi.
> [Fly61]
>      Henry Flynt. Concept art. In La Monte Young and Jackson MacLow,
> editors, An
>      Anthology. Young and MacLow, New York, 1963 (1961).
> [hun90]
>      George Maciunas und Fluxus-Editionen, Edition Hundermark,
>      Cologone, 1990.
> [Knu98]
>      Donald E. Knuth. The Art of Computer Programming. Addison-Wesley,
>      Reading, Massachusetts, 1973-1998.
> [Les00]
>      Lawrence Lessig. Code and Other Laws of Cyberspace. Basic Books,
>      New York, 2000.
> [Sha]
>      Edward A. Shanken. The house that jack built: Jack burnham's
>      concept of`software` as a metaphor of art. Leonardo Electronic
>      Almanach, 6(10). http:
>      //
> Footnotes:
> 1 Some of those writings are reconstructed as computer programs on my
> web site``Permutations'' [Cra99]
> 2 See [Les00]
> 3 Henry Flynt, Concept Art [Fly61] ``Since `concepts' are closely
> bound up with language,'' Flynt writes, ``concept art is a kind of art
> of which the material is language.''1
> 4 ibid.
> 5 [Knu98]
> 6 Among them Ted Nelson's hypertext system in its first public
> display, according to Edward A. Shanken, The House that Jack Built:
> Jack Burnham's Concept of ``Software'' as a Metaphor for Art, [Sha]
> 7 ibid.
> 8 Like, for example, John Barth's ``Frametale'' which consists of the
> Moebius strip as an infinitely recursing narrative, [Bar68], p.1-2
> 9 Or, as Adrian Ward puts it: ``Children are crafted by nature,
> software by nurture. I am the craftsman. Thus, I shall live on through
> myself'' (quoted from an E-Mail message to the ``Rhizome'' mailing
> list, March 9, 2001)

Annotated by Trevor Batten
(date unknown)