on the Interactive Art Conference on Arts Wire
Poet Jim Rosenberg is our January guest on the Interactive Art Conference. In his own words:
Jim Rosenberg began a life-long concern with non-linear poetic forms in 1966, with a series of polylinear poems called Word Nets. By 1968 this concern had evolved to an ongoing series of Diagram Poems, which continues to the present.
The diagrams began as an effort to support word clusters, by analogy to the musical concept of tone clusters; because the juxtaposition of words in a cluster disrupted syntax, an alternate channel was necessary for syntax. The diagram concept has provided a rich ground for experimentation with structural concepts not present in ordinary syntax, including null relationships, feedback loops, and interior links. This body of work includes Diagrams Series 3 and Diagrams Series 4, ad hoc circulations published on-demand by the author using a computer. Diagrams Series 4 is available on-line electronically without manufacture through The Art Com Electronic Network on The WELL, Sausalito, CA, with access via CompuServe's packet switching network, and on the World Wide Web..
Since 1988 his work has consisted of interactive poems using a Macintosh computer and HyperCard software. In this ongoing work, beginning with Intergrams, the word cluster is at last implemented as words overlaid in the same logical and physical space, with the mouse used to render individual phrases legible. This is combined with the diagram notation using hypertext links to simplify navigating the diagram. Intergrams is published by Eastgate Systems, Cambridge MA.
Jim Rosenberg was born in 1947 in Denver, Colorado, USA, and received his undergraduate education at Pomona College and his graduate education at the University of California at Berkeley -- both degrees being in mathematics. His poetry has appeared in several small magazines, including This, Tyuonyi, Interstate, Open Reading, Toothpick, Vort, and BUTTONS. He has performed his poetry at The San Francisco Poetry Center; Intersection, San Francisco; Cody's, Berkeley; St. Mark's Church in the Bowery, New York; and The Kitchen, New York. His works for magnetic tape have been performed by radio stations KPFA Berkeley, WBAI New York, and VPRO Amsterdam; and by the Stanford New Music Ensemble. "Intermittence", a poem for four simultaneous voices and conductor, has been anthologized in Scores: An Anthology of New Music, ed. Roger Johnson, Shirmer Books, 1981. He has constructed the word environments Temporary Poetry 10/73, Les Salons Vides, San Francisco, and Permanent & Temporary Poetry, The Kitchen, New York.
He has lived since 1974 in a stone house built in the late 18th century in Grindstone, PA, with his wife and co-experimenter, the painter Mary Jean Kenton. He supports himself programming computers.
Jim's Intergrams - layered poems read in part by "peeling" off layers of text - and his idea of taking hypertext into the very structure of language are radical/individual/visionary ways of looking at working with words and computers.
Jim, can you tell us a little about how you evolved (if that is the correct word here) from mathematician/computer scientist to poet? Has this background been influential in the development of your poetry?
Welcome, Jim! I'm looking forward to this...
It wasn't so much an evolution as a "co-occurrence" of both sides to my life all along. I've been writing poetry since 1962, and have been programming computers since 1964. I guess I must have started playing with mathematics on my own outside of school back around 1960 or 1961. The two sides haven't always gotten along well. In the summer of '68 I decided computers were screwing up my head for writing, and resolved that after graduating and after my summer job was done the next summer, I would throw the computer out of my life completely. I did actually do this, and went to grad school at Berkeley in math; after the summer of '69 I don't think I touched a computer for 8 years. My trajectory as a graduate student in math ground to a halt in 1974 [long story] and for a few years I tried living 100% as a poet.
Not only did this not work economically, I found there was something missing from my life. When the microcomputer revolution hit it allowed me to work out the problem of earning a livelihood while doing the artistic work that mattered, but still the two sides fight with each other occasionally. For the last few years my life has seemed pretty much in balance; though holding down a regular job leaves me less than a third of the time I'd like to *read*.
There are some pretty obvious mathematical or logical aspects to my work; I've always considered that that was "just me" -- the two sides are me, after all. (This hasn't always been easy for other people to swallow; John Cage once told me, criticizing some of my earlier diagram work, that he found it "too mathematical" -- as if this were an insult. This actually hurt quite a bit, but after a time we got past that point.)
I guess I would have to summarize by saying that my appetite for things computational and mathematical on the one hand, and for poetry on the other, have led me to purse both, since roughly the time I was in high school.
Incidentally to call me a "computer scientist" is being a bit too generous: I have absolutely no credentials in computer science (though I do have *a lot* of work experience as a working programmer.)
Thanks Jim! When/how did these mathematical and/or individual aspects - what I refer to as your "structuring" become integral to your work?
Not to deter you from answering Judy's question, because I'm very interested in what she's asking and your responses...but I'd also like to explore another track. hope you don't mind a polyphonic interview/conversation :-)
In your description of Diagrams 4, you point to the work being made "without manufacture" and published "on demand". Yet you do not term Diagrams 4 as interactive. From what I can gather from the sample of Intergrams and your subsequent work - interactivity there consists of the reader peeling off the layers?
I'm fascinated by the spatial, connection-oriented way of reading your work impels. The diagrammatic structure and the sensuality of your words (sensuality as in evoking the visual, tactile, aural) make for a reading that is both a physical or sensed experience, and one that is highly conscious - requiring activity on the part of the reader to make connections.
(& btw, it's a real treat to see your web site - & how your work has evolved over time.)
What determines, in your work, the designation of "interactive"? What role does interactivity play in the structure of your poetry.
Greetings Jim, it's great to see you here!
This is already quite fascinating, and I don't mean to add yet a third tract to this conversation, but I will anyway.
As a composer, I'm interested in how music influences your work (the mention has already been made of the analogy between your "diagram" poems and tone clusters). (One could also pose the question as, "how your work influences music.")
Am I correct in recalling that you wrote a program for John Cage to automate his "writing through" texts (I think it was called MESOLIST)?
(Well, I guess that opens up two more topics.)
More good questions :-) Glad to see you here, Doug.
> When/how did these mathematical and/or individual aspects -
> what I refer to as your "structuring" become integral to your
That's been a long process. I began experimenting with my diagram notation while still an undergraduate -- back around '68, as I remember. From there through the early '70s I worked with it off and on -- I varied a lot back then in how "experimental" my work was. I guess I became convinced somewhere about 1974 that this was at the core of what I had to do. I don't have any clear memory of just how I decided this. I simply had the sensation that was where the energy was coming from, that this was the work for me to do and no one else.
>> In your description of Diagrams 4, you point to the work being made "without manufacture" and published "on demand". Yet you do not term Diagrams 4 as interactive.
What determines, in your work, the designation of "interactive"? What role does interactivity play in the structure of your poetry? <<
The Diagram Poems were originally written to be seen on paper. The only interactivity -- in either the version you hosted on the ACEN conference of the WELL, or my Web site -- is choosing which one to read from a menu. That's not really much in the way of interactivity.
Somewhere along about '86 or so, playing with bit-mapped graphics and a mouse, I realized that software provided me a way of doing something I had wanted to do very much from the very start: word clusters -- putting words literally on top of one another. When words are put on top of one another visually, or aurally, the result often is that they interfere with one another to the point of unintelligibility. With interactive software, the words can be put atop one another and then by using the mouse, the reader can reveal individual layers one at a time, so all the words are intelligible. I don't think of this as peeling layers off so much as diving in to the simultaneity to look at the layers one at a time, though I suppose either concept would work just fine. Interactivity also provides a way to navigate the diagram syntax when it's too complex to fit in one screenful.
With _Intergrams_, you simply can't read it at all without "operating" the poem. Incidentally, I should say that the actual user interface doesn't really work the way the interface works in the extract I have on my Web site. In the actual piece, layers in what I call a simultaneity are revealed simply by moving the mouse through invisible hot-spots -- there is no mouse-clicking, though I do use conventional "click" buttons for navigating the syntax. (There simply *is* no way to do no-click hot-spots in HTML -- though I'm eager to see whether perhaps Java will let me do this.) In the Diagram Poems you can simply read a poem printed out and there's nothing to "operate"; in my interactive work, you don't get to the words without operating interactive structures that are deeply embedded in the text.
This is taken to an even further extreme in the piece I'm doing currently. The simultaneities are nested many levels deep, and there are simultaneities in some cases inside the sentence itself.
>> I'm fascinated by the spatial, connection-oriented way of reading your work impels. <<
My work has become even more and more spatial as time has gone along. There are a lot of parallels between my work and the hypertext research carried on by Cathy Marshall and her colleagues -- she's done probably the most important research in spatial hypertext. Reading her Aquanet paper I just about jumped out of my seat -- she was using a diagram notation for what are known as Toulmin structures to diagram argumentation that looked so much like my own diagram notation that it was uncanny. It seems as though there is being more attention payed to spatial hypertext lately than there used to be, which is encouraging.
>> As a composer, I'm interested in how music influences your work <<
Music influences me constantly, and I guess it's always been that way. I get tremendous amount of energy from music, and would have to say the single greatest artistic influence on my life was John Cage. Beyond the specific idea of tone clusters -> word clusters, it's hard to be specific about what exactly turns into what in terms of music influence, but it's there, as energy absorbed and recycled. These days I find myself being moved the most by composers like Ingram Marshall and Pauline Oliveros, though I like a lot of different kinds of music.
>> (One could also pose the question as, "how your work influences music.") <<
I would be thrilled to think of having influenced music somehow, but I haven't had any composers tell me this.
>> Am I correct in recalling that you wrote a program for John Cage to automate his "writing through" texts (I think it was called MESOLIST)? <<
Yes, I wrote some of the early programs in Cage's mesostic projects. Andrew Culver ended up doing most of Cage's programming.
> along. There are a lot of parallels between my work and the
> hypertext research carried on by Cathy Marshall and her
> colleagues -- she's done probably the most important research in
Thanks, Jim, for all your responses. I've been working with Cathy Marshall on a collaborative hypertext and in addition to the uncanny way that our voices blend, it has been good to work with some one whose views about what is and what is not hypertext are so open.
> so much like my own diagram notation that it was uncanny. It
> seems as though there is being more attention payed to spatial
> hypertext lately than there used to be, which is encouraging.
You and I have talked for about how difficult it is to work with experimental structures in a hypertext community where work is criticized because it is not hypertext. I've had acceptance problems with my work (like Wasting Time that Richard Gess published but no one else has been interested in) that is based on parallel streams of text. Originally I didn't call my central work (Uncle Roger, Penelope, The Yellow Bowl) hypertext but narrabase (derived from narrative database) I still like that term but found that Mark Bernstein was correct that it is better to be associated with the dominant trend. Actually talking to Cathy helped me see my work more in hypertext terms.
I was blown away when I first put mouse to your intergrams but I didn't think of the work in hypertext terms, but if you look at hypertext in terms of complete building blocks of information put together in various ways - which is how I now look at it ( but perhaps you see this differently?) than it is hypertext - spatial hypertext. Are you comfortable looking at your work in this way/entirely in this way? Does it matter to you what it is called?
And to throw in another loaded question, can you talk about language poetry and your relationship to the language poets?
>> Thanks, Jim, for all your responses. <<
Thanks all for the questions!
>> I've been working with Cathy Marshall on a collaborative hypertext and in addition to the uncanny way that our voices blend, it has been good to work with some one whose views about what is and what is not hypertext are so open.
You and I have talked for about how difficult it is to work with experimental structures in a hypertext community where work is criticized because it is not hypertext. <<
Well, maybe I'm being naive, but it seems to me there is reason for optimism. The conventional "node-link" model of hypertext is still dominant -- and the Web hasn't helped that any, though Java may turn things completely on their head in that regard. But it seems to me the hypertext research community *is* opening up to a wider concept of what constitutes hypertext. I'm encouraged, for instance, to see that there will be a whole session at the upcoming Hypertext '96 conference in March devoted to Spatial Hypertext. (With a certain very curly-haired character named yours truly one of the speakers ... :-)) Eastgate has a new product called Web Squirrel openly advertised as spatial hypertext -- the first really commercial spatial hypertext product I'm aware of, actually. (Mark Bernstein openly credits Cathy Marshall's VIKI as the origin of many of the ideas for Web Squirrel, BTW.) I can't say I've experienced the kind of "discrimination" -- if that's the right word -- you have, Judy. That may be because I'm working in poetry, and there is so much less hypertext poetry than fiction that people haven't had time for their ideas to calcify.
On the other hand, there is a definite strain in the hypertext community of folks who like to get up on a soapbox and propound all kinds of confining rules for how hypertext "ought" to be done -- I call this "fusbudgeteering". So we'll see. But I remain optimistic!
>> if you look at hypertext in terms of complete building blocks of information put together in various ways - which is how I now look at it ( but perhaps you see this differently?) than it is hypertext - spatial hypertext. Are you comfortable looking at your work in this way/entirely in this way? Does it matter to you what it is called? <<
I define hypertext as any form of text with embedded interactive structure operations. But I take a pretty broad view of what those operations might be. I'm not too thrilled with the "complete building blocks" idea. It's true that that's how a lot of hypertext has been built, but I'm much more interested in work that is built from scratch as hypertext. One of the concepts I explore in my HT'96 paper is what I call the hypertext *episode*. An episode is whatever group of activities cohere in the reader's mind as a unit. One of the ideas I'm proposing is to consider the episode as a virtual document, rather than what is known as "the lexia". For those not familiar with hypertext literary lingo, 'lexia' is a term George Landow borrowed from Barthes to describe the chunk of text at either end of a hypertext link. It is truly appalling to see how lexia-centric hypertext still is; in my paper I argue for meaning happening *through* activation of the interactive devices -- not just meaning in the lexia, but meaning in the episode. It will be interesting to see what the reaction to this concept is; so far I've gotten some pretty nice advance feedback.
My ultimate interest here is hypertext as a medium of thought. To me that *doesn't* just mean hypertext as a medium for organizing *thoughts* -- linear thoughts -- but rather as a medium in which one "thinks native" thoughts that are hypertext all the way through: hypertext extended into the fine structure of language. This is a very difficult idea, on which so far I've got very few takers.
Does it matter to me what my work is called? Not really. If someone wants to throw me out of the club of hypertext because I don't use links, so be it, but I'd rather focus on the commonalities between some of the non-link structuring methods and the more "traditional" ones. One of the things I tried to do in my HT'96 paper was to generalize some of the kinds of rhetoric Landow likes to apply to other methods, such as spatial ones. My impression is that hypertext people are increasingly receptive to this.
>> And to throw in another loaded question, can you talk about language poetry and your relationship to the language poets? <<
It certainly is loaded. I hope no one minds, but since this exact issue came up on the ht_lit mailing list some months back, I thought I would post here what I said there about it, since it pretty well says what I have to say on this issue. This was posted to ht_lit (Hypertext Literature mailing list) back in May.
>> Jim has long been associated with the L=A=N=G=U=A=G=E Poets <<
There is nothing in the whole realm of poetry that gives me more anguish than this question of what is my relationship to these people. The statement above is true in the literal sense that I've known many of the L=po folk since before there was an L=mag; we *do* share some concerns, but there are also some points of major difference. I have said it before and will say it again, even though it now sounds somewhat silly (I am not so naive as not to recognize that 'language poetry' as a term is an accomplished fact of many years' standing): the term 'language poetry' is odious, is hurtful; it hijacks the word language, which belongs to *all* poets. Poetry by those not of the L=po persuasion is made out of *what*?!
That having been said: I like the idea of thinking of language as one's material as visual artists might think of their material: something that can be manipulated, rearranged, put through *process* perhaps. I *am* interested in "language itself", I do believe my work raises questions about language, but those are *not* what the work is "about".
>> which indeed shake up the language and to realign meaning radically.. <<
As usual, Michael puts it beautifully -- certainly I would drink to this.
Where the L=po group and I part company -- drastically -- is on the issue of poetics. The L=mag established an atmosphere of poetry and poetics in close confines, on the same ground, in the same idiom even. While I can stand back and look at Charles Bernstein's formulation "Poetics is simply a continuation of poetry by other means" and whistle to myself "slick move!" I find myself taking violent objection. If you say poetics is simply a *continuation* of poetry by other means, there is a natural implication that if you don't do poetics you have cut something off, have stopped a natural continuation. This is wrong, wrong to the point of potential harm. Perhaps for Charles Bernstein ceasing to do poetics would feel like the cutting off of a natural flow that comes directly out of the work, but what is harmful here is the presumption that it applies to everyone: there is a rank prejudice rolling around in the L=po world which says: "thou shalt DO poetics -- lest thou not be taken seriously as a poet." This is just plain rubbish.
There are those who are conscientious objectors on the field of poetics, who feel -- passionately -- that the work must speak for itself, who simply don't do poetics. While I do do poetics from time to time -- in measured doses, hopefully -- we must absolutely respect the refusal to do poetics as just as valid: it is quite literally true that in the end it is the poetry that matters; if the poetry works without the aid of poetics, there is nothing "missing" by its absence, no fault attaching to the poet, no excised connection to be looking for or wished for, and absolutely no lack of seriousness.
There is an analogy from biology that I find useful as a metaphor for the artistic process that may help explain how I feel about poetics. We all are so used to navigating by senses that operate via direct lines that we can't even imagine what it would be like to function in a world where the primary sense is chemical. Some organisms that orient to chemical gradients use a mechanism called klinokinesis. (I hope I'm getting this right ...) These animals are constantly *turning*, and the rate of turning is adjusted based on whether the turn took them to more favorable or less favorable conditions. It works well for orienting to chemical gradients, and even though these animals can't "see" the source of their food, by the laws of probability the turning is controlled in such a way that they end up at the points of maximum concentration of the chemical to which they're orienting.
This says a lot to me about the artistic process. One *is* constantly turning, and the important thing is that fine-grain sense of feedback: this worked, that didn't, this feels right, that doesn't. One can't always directly *see* hulking above the landscape the exact artistic endpoint, but ultimately the feedback from turnings enables that journey that somehow gets you to the point of maximum energy. Poetics, on the other hand, runs the risk of erecting a line-of-sight structure that sticks out above the landscape. There's nothing wrong with that, of course, but woe unto you if in the process you manage to blast those sensory nerve hairs that make the klinokinesis work. (And shame unto you if you teach young people that the klinokinesis process doesn't matter ...)
Michael writes so eloquently and often about the *topographies* of writing: having gone a journey by klinokinesis, one *may* be able to see what the topography *was* and explain it in a way that is useful to others; I suppose I tend to think of poetics as being a more retrospective activity than a prospective one. The risk is that in spite of one's intentions, the theory ends up speaking with a louder voice than the klinokinesis. Sometimes you just have to tell the theory to shut up. The non-linear work that I did prior to _Intergrams_ was a series of diagram poems going back a long way. In between _Diagrams Series 4_ and _Intergrams_ I did a whole series of good ole fashioned linear poems. I didn't really "decide" to do that, and can't really justify it in any theoretical way -- it just happened. I did a couple of linear pieces and just kept coming back and coming back because something said that was a turn I had to make. The piece I'm working on now will probably *not* use my diagram notation. Again, this just "happened", it might go against some of the neat theory I've got -- but when the microturn says "yes, this way" you have to listen.
I suppose what I would like to keep from L=po is the inquisitiveness and energy about language as material, as process, but I don't have much affection for the rhetoric.
Thanks Jim - I'm looking forward to reading this offline. AS an aside, I was over at Poets & Writers this afternoon and I got the url for the Electric Poetry Center at SUNY Buffalo. It's http://wings.buffalo.edu/epc but I haven't looked at it yet.
>I define hypertext as any form of text with embedded interactive
hmmm- this is a good definition that fits both your work and work with explicit links (Joyce, Moulthrop, Guyer) , but in that it defines hypertext in terms of navigation and structure instead of in terms of writing (or perhaps Bolter's "writing with places"), it is also a somewhat limited definition.
I would argue that using complete (screensized) building blocks of text (lexias) written with links that are implicit but not necessarily defined to the reader is also hypertext. With Forward Anywhere (the work that Cathy Marshall and I did) we were thinking in hypertext terms as we wrote in that we mentally choose something in the other's preceding text to link to. Quoting Cathy from our paper for the Wired Women book:
"But during our second meeting, we decided on a course that avoided unnecessary invention -- to exchange the real remembered substance of our lives. It is odd how the links arose quite naturally all along within this constraint, although I stewed at great length over the first screen I sent Judy -- too glib, too vague, I thought, to be evocative. The "lived lives" constraint turned out to be a positive force in the process, shaping the work in an organic way.
We agreed, at our meeting in Berkeley, to define structure later in the project, after we had amassed content by email. The links were to be left more or less implicit in our exchange (although, in practice associations often found their way into the Subject line).
In retrospect, email seems like a naturally hypertextual form, with its splitting and merging threads of conversation, its subjects that recur and re-emerge, and its tendency to discourage linearity......"
(Malloy and Marshall, Closure was never a goal of this piece.)
When we structured the work, we choose not to embed the links within the text and were somewhat uncomfortable with the semi-explicit "forward", portion of our interface (I had used this method to structure "Scibe") because the work *itself* seemed to us to be so clearly hypertext (interactive perhaps in its collaborative nature) in the process in which it was written. Quoting Cathy again from the same source:
"Besides adding the gathering function described earlier, we also decided to include a random function that brings a reader to a screen selected at random from our entire collection. The random function addresses the high interconnectivity just as surely as using a large number of explicit links, since the effect is in some ways quite similar -- a reader can get to any screen from any screen. It is the forward and backward functions -- and other explicit links -- that have given us pause. It is perhaps the sum of the experiences of past screens, the cumulative mystery, which has lead us to write the next. "
Turning this back to your work, I'm interested in how the Intergram structures evolved. How do your writing processes/your content relate to your structures?
Oh and thanks for the excellent response to the language poetry question.
>> hmmm- this is a good definition that fits both your work and work with explicit links (Joyce, Moulthrop, Guyer) , but in that it defines hypertext in terms of navigation and structure instead of in terms of writing (or perhaps Bolter's "writing with places"), it is also a somewhat limited definition. <<
Before getting into the substance of Judy's argument, there are some things I should say. I'm a pluralist, and get quite upset with people who think there is only One Way. On a certain other mailing list (that I'm not supposed to mention in cyberpublic) with lots of poets participating, they seemed pleasantly surprised when I openly stated my support for non-computer methods (like physical transparencies, etc.) I'm on record as saying that I don't think we hypertext folk are doing ourselves any good by what I describe as unguided missiles talking about "death of the book", "the late age of print", etc. In doing what I do, the last thing I want is for anyone to feel that I'm trying to *preclude* some other way of working. As I like to say, there's one house of poetry with room enough for everybody.
That having been said:
>> I would argue that using complete (screensized) building blocks of text (lexias) written with links that are implicit but not necessarily defined to the reader is also hypertext. <<
I don't know whether I buy this. Many people have argued that there are "implicit links" to a lot of different kinds of print literature, also. It's a bit painful: as with any definition, we either have to stop using the term, and admit it's useless, or we have to accept that it does place a boundary somewhere. I suppose I wouldn't mind having to give up the word 'hypertext' -- though when I'm described as a hypertext poet I don't disown that. Cage taught us to think in very large terms: The domain of music is anything which can be heard, the domain of visual art is anything that can be seen. And of course, there are times when even these terms seem too confining. To me, the domain of poetry is all of language. So yes, certainly, I'm interested in an inclusiveness that embraces "ready-made" lexia with implicit links. On the other hand, I'm not sure whether I'd call it hypertext. Maybe we need a better term. Cathy Marshall has told me that there are some well known hypertext folk who give her the impression all the time that deep down they don't consider what she does as hypertext; in that case I'd be outside the fold too.
At bottom, as a "card-carrying experimentalist", I put the questions first: *What happens if* -- what happens if I deeply interlace explicit structure operations into the fine structure of language. On the other hand, I support completely the idea of starting with the words and having the structure "migrate its way" into the words: that's the way I've been doing my work for quite a while now, actually. Everyone has to follow their own path to how the energy works; each experimentalist has to formulate her own questions. Implicit links are interesting! Maybe we need a new term that embraces everything from hypertext narrowly construed as the [notorious] node-link model, to print text that invites non-linear exploration by implicit associations. What we're really talking about here is a kind of "opentext".
>> Turning this back to your work, I'm interested in how the Intergram structures evolved. How do your writing processes/your content relate to your structures? <<
I've developed what I call my "semantic method" over many many years; in some ways it works independently of the structures. I use what composers call "precomposition". That is, the piece is composed in layers of activity, and each but the final layer affects the entirety of the finished piece. I maintain what I call "reservoirs" (Jackson Mac Low calls them "vocabularies") which are precompositional groupings of phrases. At each step I take the existing generation of reservoir, chop it up, permute it using chance operations, and use the resulting "prompt sheet" to write -- by hand, so to speak -- the next-layer reservoir. At the end of this process I am writing finished phrases from the last-layer reservoir. I will typically do quite a number of pieces from a given reservoir.
Somewhere about the middle of Intergrams I began doing an "edit phase" that has become very important to me. When I have the first draft of finished phrases, I cull these. A few -- I would guess typically no more than about 10% -- are good enough to go into the finished work mostly intact, with just minor editing. The rest go into a "metamorphosis soup". (When those insects that undergo complete metamorphosis enter the pupal phase, their bodies literally dissolve into a soup of cells. Many of these cells migrate to completely different places in what will be the adult insect, and a whole new kind of organism is assembled from the cells. That's not unlike how my edit phase works.) I pick out the pieces that still work, and then pull in words or pieces from "sacrificed" phrases reassembling new phrases. I do "cheat" occasionally -- putting in a word that didn't come from the first finished phrase draft -- but this is pretty rare.
From the result of all this, I pick the final phrases for the finished work. My wastage percentage here has gone down quite a bit; it used to be that I would throw out about 2/3 of the final phrases, but in the work I'm doing now, _The Barrier Frames_, the percentage of kept phrases is pretty high.
At this stage I move the phrases about letting them attract one another into clusters, with the structure emerging from the words. Composition of the words has normally taken about 85% of the elapsed time of making the work.
With _The Barrier Frames_ things have worked out somewhat differently. I composed simultaneity structures in the small as I went, then redid them using the same edit phase mechanism; in the large the clusters emerged from the finished words, as before. Also in this piece, the "assembly" phase has been taking me *months* -- it's excruciating!
Some of these methods are compromises based on the available tools. I simply don't have an authoring system that will let me make reservoirs out of pieces with the interactive structure operations built in, which I can then "plug in" elsewhere. I have dreams of an authoring system that will let me do this, but don't have time to program this myself. (When I was an undergraduate, I knew a guitar maker who lived in the town where the college was; I would hang out there occasionally. I was naive enough back then that when he said he couldn't afford his own guitars, I was shocked. I sure understand it now: I earn my living as a programmer, but as far as using my programming skills for large projects involving my writing, I can't afford myself!)
>Everyone has to follow their own path to how the energy works; each
>experimentalist has to formulate her own questions.
Yes! This is by far the most important thing and I agree 100% Every writer, every artist has his or her own way. That's one of the reasons print is not going to disappear. This week Michael Joyce's phrase "print stays itself; electronic text replaces itself." came to mind. I realized that it is also an apt definition of your work, (Do you think so?) but it is nothing like my way of thinking.
> Maybe we need a new term that embraces everything from hypertext
> narrowly construed as the [notorious] node-link model, to print
> text that invites non-linear exploration by implicit
>associations. What we're really talking about here is a kind of
For me the ability to manipulate the text is very important. That manipulation is what computers make possible. I'm less interested in so-called print hypertext - perhaps because I spent some many years trying to make nonsequential text work in print so I know first hand that however you structure it, it will be read sequentially. I'm rereading Hopscotch (Anna gave it to me for Christmas!) and I don't skip around in it. I read it from cover to cover. It is an excellent book, but I feel no kin with its structure.
>On the other hand, I support completely the idea of
>starting with the words and having the structure "migrate its
>way" into the words: that's the way I've been doing my work for
>quite a while now, actually.
Working on the Yellow Bowl has been an epic interface battle - symbolic perhaps of the difficulties of attempting to combine sequential and nonsequential text and of the primary importance of words and intention. YB was designed (I also use a form of precomposition that involves large charts on my walls) with what I conceived of as three streams of text - two were sequential (the stories that the narrator told her child) and one was pseudorandom (the narrators thoughts/memories that she distorted to shape these stories -designed to appear in changing but natural memory patterns) The reader was supposed to move in and out of the stories and the narrator's memory bank. I used peripheral "frozen" links. ("buttons" - I hate that word)
In installation situations, I discovered that people didn't bother to get into Grace's mind. They stayed in the stories til they got to the end. So, I experimented with arbitrarily throwing them into the memory bank at certain intervals and tried it again. They got out as soon as they figured out how- the human desire perhaps to know the end of the story. But in this work, the stories by themself were not important. It was how they were shaped from the narrators experience (and how parents communicate experience to their children) that was backbone of this work.
So had to change the interface. (although the underlying structure remains the same) The work was already written to links (implicit to the reader but explicitly written to) so it was easy to highlight them using a "linkplot" strategy with dense linking between the three "streams" of text that guided the reader with links that sometimes produce the random material and sometimes continue the story. Currently, YB is more "authorial" than I would like, but I think that it now "works" in a fairly seamless way. However, I wouldn't make changes because of how readers experienced the work in every one of my works. In this work, the reading pattern was central to my aims.
Poetry is different than narrative. The roots of your work, the writing of your work are very different from the roots and the writing of my work. Nevertheless, I'd be interested in how you have observed readers interacting with the Diagrams. Has reader experience has shaped your work in any way? Is that is important to you?
>what happens if I deeply interlace explicit structure operations into
>the fine structure of language.
I would like to get more deeply into what you mean by this. Feel free to plunge in if you want. Meanwhile I am hoping some of the musicians on Arts Wire will to address your endlessly fascinating composition methods.
oh and I meant interacting with the Intergrams.
>> Poetry is different than narrative. <<
Right. I don't have "the" story to which I have to hold any allegiance.
I suppose, as a reader, I'm something of an oddball. I think to most people the essence of a good book is one that you can't put down, you have to keep reading to find out what happens. To me, the essence of a good book is that you *have to* put it down because the imagery is so rich, is resonating so fully, that you have to put it aside and let it ring for a while. Resonance is clearly a whole different animal than narration.
>> Nevertheless, I'd be interested in how you have observed readers interacting with the Intergrams. Has reader experience has shaped your work in any way? Is that is important to you? << [Correction Diagrams -> Intergrams interposed ...]
I suppose perhaps my answer here will bother some people, but no, I haven't "studied" what readers do with _Intergrams_ or _Diffractions through_. I get a curious feeling when I'm present and someone is operating my work; I feel as though I'm in danger of eavesdropping on what should be a private moment between the reader and the work, and always have a *very* strong inclination to want to leave, to leave them alone with the words. I can see where some people would feel that part of the job is studying the ergonomics of one's own creations, but I see it differently. When you get past all of the technology and the buzzwords, what it's all about is nothing more complicated than simply:
bringing words together
and ultimately, as an artist, I have to rely on my own convictions and intuitions of what's right and what's not right. I like your idea of studying what readers do to try to "throw them out of the narration channel" -- maybe you wouldn't like how I've put that -- but I've never felt impelled to do anything along these lines.
Of course as I'm making new work, there is a handful of people whose opinion matters a lot to me; I will bug them about whether things "work" or not, and may have an occasional precise question or two about whether specific interface features work the way I had in mind, but I don't really "study" what readers do with my work.
I'm really eager to see what you're doing in the Yellow Bowl.
>>> what happens if I deeply interlace explicit structure operations into the fine structure of language.
>>I would like to get more deeply into what you mean by this. Feel free to plunge in if you want. <<
This is a large subject, which I've been grappling with for a very long time and about which I care very deeply. The rudiments of my diagram notation first came to me somewhere along about 1968; at first I simply conceived of it as a device for making poetry, but in the summer of '69 I had a flash one night at about 2 in the morning and realized that "language itself" could be based on these concepts. In all the years since I've been waging an ongoing battle with myself to try to get the ideas more and more explainable, but it is a really difficult concept. Eduardo Kac is guest-editing an issue of _Visible Language_ that will have papers in it by himself, the British computer poet John Cayley -- whose work I admire very much and several others; I have a paper coming out in that issue called "The Interactive Diagram Sentence: Hypertext as a Medium of Thought" that goes as far as I've been able to get lately in explaining this. I wish I could post that entire article, but I don't know whether I can do that. Most people approach hypertext from what I would call the outside in. That is, they start with whole documents -- typically documents written "pre hypertext", with no interactive structure at all. They then begin to impose more and more interactive structure on the document; perhaps section navigation, table of contents, etc. are handled through links, a few key words or topics are made into link anchors, then gradually more and more links are made, and so on. The focus is still "the document".
By having started with my Diagram Poems, I was going about it exactly in the opposite way. We normally think of hypertext as software, or as the electronic documents created using software, but I think of hypertext as a virtual *diagram*. You certainly get a diagram in a conventional node-link hypertext if you diagram the link network. The Diagram Poems were a sort of proto hypertext "from the inside out". The diagram structure forms the very basis of the syntax; rather than starting with linear documents and using "hyper" operations to connect whole documents, or sections of documents, the very sentence itself is constructed using these methods.
It's very important in understanding how this works to keep in mind the and/or distinction, or what I've called conjunctive vs. the more typical disjunctive hypertext. In the usual approach to hypertext, the reader has a choice among *alternatives* in choosing which link to follow. If lexia X has links A, B, C, the reader can choose A *or* B *or* C. I've been arguing that it also makes sense to build hypertext based on "and" rather than "or". In logic, "and" is a conjunction, "or" is a disjunction, so I call this conjunctive hypertext rather than disjunctive hypertext. (The paper where I introduced all of this is Navigating Nowhere / Hypertext Infrawhere) Conventional syntax is conjunctive: if we say a sentence is composed of a noun phrase and a verb phrase, the noun phrase and verb phrase are hardly *alternatives* -- neither one is "optional", they're both there, you have to have the noun phrase *and* the verb phrase to get the sentence.
Sentences are also a kind of virtual diagram. You could say that syntax has been nature's way of allowing us to fit very complicated structure relationships into a word stream that is constrained by two important limitations: (1) Throughout most of the time speech has been existence, speaker and listener have had to be synchronized in approximately real time. (2) The word stream itself has no storage; storage has to be achieved by speaker and listener. You could say that syntax is a way of encoding *storage cues* into the message so that structural relationships among distant parts of the word stream can be constructed in the face of these constraints. But writing does not have either of these constraints. There is no time synchronization constraint at all with a written document, and the document contains its own storage. In writing we are free to indicate structural relationships *directly*.
This is what I began doing in my diagrams. Of course the problem with a diagram is how do you get it to scale into an immense number of elements. This is where hypertext comes into the picture. Hypertext lets you build a virtual diagram of unlimited size. These hypertextual operations can carry the same structural relationships as syntax -- *inside* the sentence Meaning is not just a function of the lexia, but happens has we move *through* the links, at the level of what I call the *episode* in a paper I'm giving at the Hypertext '96 conference in March.
The issue here really is hypertext as a medium of thought. By this I don't mean hypertext as a medium for organizing *thoughts* -- linear thoughts which are not themselves hypertexts, but rather as a medium for "thinking native" in hypertext, where the *individual thought* is a hypertext, "all the way down". In the _Visible Language_ article I raised the question of what would it mean to construct a *multiuser sentence*. That raises the issue of hypertext inside the sentence in a context where I don't see how to do it any other way.
How do we construct a medium where *an individual thought* can be a multiuser construct?
Thanks Jim! I'm really enjoying this conversation and am looking forward to reading and responding offline. Thanks again for taking the time to visit us.
Jim, there's a lot you've said all thru this topic that has sparked much thinking. I suppose by your definition of a good book, this is a good topic, because I want to absorb it a while before tossing a response back at you.
About the levels of interactivity in Diagrams & Intergrams. Your answer seemed so obvious when you said it. It wasn't until later that I remembered what may now be an art expression rendered technologically obsolete - that the choosing of an online work from a menu was in effect causing it to appear. And we considered it more interactive than a static physical art work because the work wasn't there until the viewer called it into being. Intergrams carries reader choice into the poems. It is several steps deeper in, and is certainly more flexible than the interactivity of Diagrams. The reader is not physically altering your work - that is, each reader will see the same work, albeit through different paths. Is the interactivity in the ability for the reader to select a path through the work?
A short question about your last post...have you explored whether hypertext as a medium of thought exists in other languages?
>>To me, the essence of a good book is that you *have to* put it down because the imagery is so rich, is resonating so fully, that you have to put it aside and let it ring for a while. Resonance is clearly a whole different animal than narration.<<
Resonance is a vary good word that is applicable to hyperfiction as well as poetry. I think in terms of small very rich pieces that the reader assembles in different ways but that stand on their own because of what you call their "resonance".
assembles I like the way you put this:
>>I've been arguing that it also makes sense to build hypertext based on
>"and" rather than "or".
I tried this using a frozen interface with selected keywords in the original second file of Uncle Roger because coming from working with databases I was accustomed to using "and" as well as "or". In this work (The Blue Notebook) the reader could make a path by combining several words. The problem was that reader "searches" often produced minimum results or no results. In a narrative, this interrupts the flow. (Although I can now see instances where this interruption could be a part of the narrative strategy, and in addition I am now more experienced in writing in tandem with the structure) Actually I think it would work much better with poetry and in particular I would very much like to see how you would apply this to your work.
>you have to have the noun phrase *and* the verb phrase to get the
And I never thought of it in this language structure way. This sounds very interesting.
>>Meaning is not just a function of the lexia, but happens has we move *through* the links, at the level of what I call the *episode* in a paper I'm giving at the Hypertext '96 conference in March. <<
Yeah - I tried conventional linking with l0ve0ne and found that in the on the click-happy web, it was being read in this way - click click click. It was a chain of linked phrases that stuck in the reader's mind and to a certain extent I began to take this into account as I wrote.
>>The issue here really is hypertext as a medium of thought. By this I don't mean hypertext as a medium for organizing *thoughts* -- linear thoughts which are not themselves hypertexts, but rather as a medium for "thinking native" in hypertext, where the *individual thought* is a hypertext, "all the way down". In the _Visible Language_ article I raised the question of what would it mean to construct a *multiuser sentence*. That raises the issue of hypertext inside the sentence in a context where I don't see how to do it any other way. <<
I have been in therapy because of post traumatic stress syndrome and have had to confront how my own thinking about life/thought processes have merged with hypertextual writing. This is slightly different from what you are talking about, and I'm not quite ready to put it in words. Nevertheless, it is very interesting.
>How do we construct a medium where *an individual thought* can be
>a multiuser construct?
Can you talk a little more about this?
>> ... that the choosing of an online work from a menu was in effect causing it to appear. And we considered it more interactive than a static physical art work because the work wasn't there until the viewer called it into being. <<
Oh dear. I didn't think of what we were doing on ACEN in these terms *at all*! It simply isn't true that the work "wasn't there" until the viewer "called it into being". Maybe it's my programming background -- the fact that I know what's really happening in the system -- but I simply can't think of it that way. All those menu choices that the reader hasn't made *are* there: they are files in the system. The reader's choice doesn't call them into being, menu choice is just a more convenient method of saying "cat /home/j/e/jer/diags/17", but the user doesn't "create" that file. The user asks to have the file *transmitted*. In a way, it's no different than special-ordering a book and asking to have the book *transported*. Of course the latter is a manufactured object, but in the former case there is no manufacture unless the user prints it on her own machine.
>> The reader is not physically altering your work - that is, each reader will see the same work, albeit through different paths. <<
Yes: this is the distinction Michael Joyce calls Constructive vs. Exploratory hypertext. There are those in the hypertext community who really push for constructive work, where what the reader does actually changes the work that *others* will see. The work is a genuine collaboration between readers and writer; in effect all are writers. There is a lot of hypertext rhetoric about reader-as-writer, to the extreme that in some cases one gets the idea that works that are only readable without the reader being able to change them are somehow "inferior".
This sentiment causes me a lot of anguish. I like the idea of constructive hypertext; comparable to what Judy was saying about the non-linear aspects of all writing, I think of meaning as a constructive act, even if the reader is not literally altering the work, and am pleased to see experimenters pushing the envelope by making works where the reader is a participant in every sense of the word. From time to time I think that I should be working this way too, but when I ask myself, "OK, *what*?" -- what would I like to do -- I don't come up with anything. In some ways I guess I'm a loner at heart. There are many people these days who push very hard a collaboration ethic that seems to imply that collaborations are inherently "better" than individual works; I think we need *both*. Collaborations haven't really happened much in my artistic life, it's just how things have worked out.
>> Is the interactivity in the ability for the reader to select a path through the work? <<
Sort of -- though I'd step back a bit from the word 'path', since that implies a choice among alternatives (or vs. and again ...) I think of the interactivity this way: I've presented the reader with word objects, where intelligibility can only be recovered by *doing things*, by "operating" the objects.
>> A short question about your last post...have you explored whether hypertext as a medium of thought exists in other languages? <<
That's a really interesting question. I've often thought that a language like Chinese would be better suited to non-linear writing than English; because the word is so spatially compact, it provides more room to maneuver, spatially. It turns out there is a tradition of non-linear poetry in China that goes back 1500 years!! There is a work known as a Poem Block, attributed to Su Hui from the 4th century A.D., where the characters are arranged in a rectangle that can be read in practically any way; *hundreds* of such readings have been constructed from her block. I think of this from time to time when I'm in danger getting too uppity about how wild and new what we're doing is. It's pretty humbling to think the Chinese were doing this 1500 years ago!
>> I think in terms of small very rich pieces that the reader assembles in different ways but that stand on their own because of what you call their "resonance". <<
Yes, absolutely: assembly is a *wonderful* word!
>> > How do we construct a medium where *an individual thought* can be a multiuser construct? <
Can you talk a little more about this? <<
I guess it's a bit ironic, given what I said in the previous message about collaboration ...
We tend to think of the sentence as a "single voice". Even when a work has multiple authors, the usual method is to strive for a "collective voice", so that each sentence sounds as though it has a single (albeit collective) author. In a diagram syntax, a single sentence can be assembled by several people; the contribution of different people could be indicated say by color. I suppose in a way this isn't a lot different than a hypertext in which different links and anchors might have different -- and indicated -- authors, though it's carried out at a very fine scale, inside the sentence.
Almost everyone these days seems to understand how huge the influence of networks is on our lives: they are everywhere, they pervade all facets of our lives. Let me pose a very naive- sounding question: How does a single mind apprehend a network? Might it not help to *become* the network? Might it not help to "think native" in a network language, in a language in which an individual thought *is* a network?
This is really at the heart of what I'm talking about when I speak about taking hypertext into the fine structure of language.
Thanks, Jim. It has been wonderful having you here. Is there anything that you particularly want to talk about?
I've enjoyed this immensely! There simply is no way to sum up the issues we've been talking about; one simply has to be clear that it's the questions that matter, that this is drastically unfinished business, and hope everyone will keep on experimenting away. When I first joined ht_lit -- and one of the first questions here! -- the issue was defining hypertext. Actually, my favorite answer to that *really* is: "We're working on it ..."
That seems like a pretty good note to end on. Thanks to you, Judy and Anna for a wonderful discussion.
Transcript of A Conversation with Jim Rosenberg, Item 74, Interactive Art Conference, Arts Wire.
Posted to the Web with participants' permission. For information about republishing, please contact Anna Couey and Judy Malloy.
Conversations with Artists