Hanns Holger Rutz


Writing (about) Writing Machines

across the different writings and different types of writing, there is a moment of selfsame authorship, voluntary or not. i have noticed this before, that the way i walk towards and away from a “text” follows similiar patterns, whether i write a piece of electroacoustic music, a sound installation, a software, this text. if all these writings are framed in terms of the investment made (long before the cut is made to begin a “new” writing), it is not surprising. strategies have been developed, some successfully tested, frustrations are avoided (or sometimes desired). a similar structure of mind and body is employed when writing different things. this not just a question of “style”. i have stopped writing the previous paragraph before the weekend, i interrupted, some threads got lost, some energy was recharged, some sleep and dreams came inbetween. a rhythmic punctuation has been inserted, as a consequence and not as a willful gesture.

the detachment between writing and reading, in the sense that the latter performed by an audience does not clearly feed back into the former, may be perceived as problematic. i am not talking about participatory art which i am sceptical of. but rather a solipsism that is allowed to enter the artistic process, which seems anachronistic. the employment of computational writing partly compensates for this, as the investment in the computational machinery produces agency that is not simply a projection of the sole mind and body, but includes history, micro-history, and contemporaneity of everything participating in the computation.

would could it mean to write with others? three things i found interesting as building blocks: first, Sara Ahmed's figure of “simultaneous arrival” at a scene, where different actors and elements emerge from a background that is both a figurative background (as in gestalt) and a historic background. and where the arrival is not just pure chance, but the coincidence coincides with an intention to become near. the movement of the arrival is what is interesting, not a “plan” to implement upon arrival. the simultaneity here highlights the nature of coincidence, everyone and everything maintains their proper time, there is no request to establish an order of cause and effect. second, Jean-Luc Nancy's idea of individuals as a multitude of “inassimilable” origins (strange to each other), instead of a force of collectivisation. meaning then is not the common medium in which collectivity is immersed, but it arises in the act of sharing in constant circulation and in a curious spacing and distance to one another without being totalised. third, Isabelle Stengers' figure of “relaying” and “relayers” (derived from D+G). in making a relay, thinking-with-others could become writing-with-others, perhaps. the gesture of the gift, which points to the hand held out for someone else to take an element and a concern and shift it, make it “my concern”, where being faithful to what is being given does not mean to engage in commentary. something is truly moved, “Your perceptions are yours, or rather are yours to work with”.


{persons: [Sara Ahmed, Jean-Luc Nancy, Isabelle Stengers], keywords: [writing, rewriting, authorship, mind, body, style, gesture, reading, audience, solipsism, process, computational, agency, history, computation, simultaneous, arrival, simultaneity, background, intention, coincidence, movement, time, space, sharing, relaying]}



it is not (just) imprint – technologically transferring bytes onto a storage medium.

it involves a program, what is written is being computed.

a program is not (just) running by itself, it has material consequences, a trace is produced by the writing.

provisions have been made for the computer to write.

writing takes time, even if what is to be written has already been formulated.

I don’t mean to write “on” a computer, a typewriter (German ‘Schreibmaschine’, literally Writing Machine). Although that can be interwoven with computational writing.

Surrealists experimented with automatic writing. Becoming machinic, shifting from premeditation to movement in itself. But I don't mean to “channel” as a “medium” (imprint).

For example, give a time span, and write within that time span, stop when the time span has elapsed. Then move on to the next question.

{keywords: [imprint, writing, storage, bytes, program, computation, traces, provision, computer, automation, premeditation, movement, medium]}

the cliché of the artist is excessive writing, the clićhe of the scientist is excessive reading (before adding a modest contribution, read ahead by peers). an idea of score is to separate writing and reading, there can be a spatio-temporal gap between the two. a program allows for a spatio-temporal gap between composing and performing. (although writing the program is also a performance).

the gap also allows to read even when I'm weak. I can be too weak to write. (when I wrote, I had sufficient energy). reading allows integration; one can read in one moment what was written over stretches of time. reading allows disintegration, one can come back and read anew and differently.

the technical terms reading/writing suggests a clear complementarity between the two. in reality, I suspect that these are different qualifications of a common underlying operation. as if thinking of a wax cylinder and a needle that can both incise the cylinder or scan what has already been written, depending on the “impedance”.

so perhaps the focus on “writing” is concealing other relations that are more important? going back to the ‘diaphragm’ or ‘score’, as a coupling between different entities that write and perform; one could instead look at coupling/decoupling. there is a tendency of technical components to be coupled and synchronised, establishing cause and effect (perhaps mutual). network metaphors are abundant. I like to think of a situation of simultaneity, where things are done together but at the same time in independence. one is not subject to a technological regime of compatibility. a situation is co-determined by what happens simultaneous, without a request to take over an intention.


{keywords: [writing, artist, scientist, reading, score, space, time, program, composing, performance, integration, disintegration, wax, cylinder, coupling, synchronization, simultaneity, independence, technology]}

not because I find authorship a very interesting question; it's more about understanding how agency is composed. as Hayles notes, not just the inscription is part of writing, but also the surface of inscription.

when writing an algorithm that produces a certain behaviour of rendering future sounds and sound structures based on conditions (environmental conditions?), this question becomes interesting.

we thought about algorithmic agency now for over three years, it is not something that has a simple answer (otherwise it would have been disappointing for such a long running project). I always thought that Karen Barad's idea of agency emerging from the intra-action of researcher and apparatus was a concise description.

“I write on a typewriter … and my machine … is the biggest influence on my work … The type-face is a standard pica; if it were another style I’d write (subtly) different poems. And when a ribbon gets dull my poems I’m sure change” (Aram Saroyan 1969)

I think the most frequently asked question is why I chose Petri dishes for this work. a short answer is: I like the sound of glass, even of the tiny glass surface of a Petri dish, which is very different from the resonances of a larger glass. sand to glass, something is becoming transparent. Transparent is not invisible. Victor Jaschke asked about this transparency when making the documentation of the piece. glass points to its shattering. in 2008, we created sound installations based on recordings of scientific spaces often inaccessible to the general public. in the archive of the phyletic museum in Jena, Germany, I recorded small glass test tubes containing the specimen, which are still one of my favourite sound recordings. handable (handhabbare) containers as a means to create taxonomies (neat separation).

question–answer; is this how artistic research is supposed to operate? “tell me something about your piece. why …” Politicians acquire a skill of answering to questions never asked. John Cage famously had answers prepared before questions were asked (by whom, was it Daniel Charles? I don't remember). Donna Haraway coined the concept of response-ability. Perhaps one could also add account-ability, the ability to give an account. Piecing together an account.


{persons: [N. Katherine Hayles, Karen Barad, Aram Saroyan, Victor Jaschke, John Cage, Donna Haraway], keywords: [authorship, agency, inscription, surface, emergence, intra-action, researcher, apparatus, writing, machine, typing, petri, glass, resonance, questions, answer]}

Nayarí Castillo – Natural Palimpsest (2009)

{kind: caption, group: nat-palim}

literacy; and when is something deemed important enough to be written down.


{keywords: [writing, literacy]}

A “dust” process is added, where from time to time a table makes a “silent” round, playing a low volume dust crackling sound instead of the sound phrases, making it possible to focus the ear on the other tables, but also highlighting the acoustic properties of the Petri dishes themselves, which you can hear shift from node to node as the wires that transmit the sounds have slightly differening positions. Each table has different dust frequencies, making a correspondence with the granulation of the graphite powder.
{keywords: [dust, petri, glass, crackling, table, graphite, powder]}

Collaborative exhibition “Writing Machines” (Nayarí Castillo & Hanns Holger Rutz, 2012) – there are individual niches, where communication happens across space, but also pieces in which elements of both artists come together (the mirrors in “Voice Trap” reflecting the brown paper piezo speakers on the ceiling; sound and visual components forming an ensemble) or are even exchanged (Nayarí making a sound composition for Hanns' red transducer-glass panels).

Week of gathering at IEM CUBE in July 2020 for the development of the sound installation “Through Segments” – taking a pause.

temporal antecedence. pro-gram, write before the writing. but actually iteration: write, run, stop, write, run, stop. sometimes (re)writing while the writing is running.

like the typical pidgeon sound. one is tricked into perceiving the end as the beginning of the phrase.

writing as doing; it is more interesting to observe the computer writing (as operation) than to observe what is being written. a “piece” is just to say, we remain with the observation, and when we return to writing then we turn to “another” piece. so “pro-gram” is always meant to introduce another future writing.

creating the conditions so that something can happen.

one sometimes employs the terms ‘meta language’, ‘object language’ to denote a nesting: natural language, the scala programming language, the sound processes computer music language. but is not a descent, one can always make the opposite movement and return.

indeed language may be the wrong entry. writing does not require language. how to write in artistic research without resorting to language? (music/sound is not a language)


{keywords: [program, time, writing, iteration, rewriting, doing, performance, observer, observation, time, future, past, language, movement]}


private val identifier = "database-query-impl"

val matchDurationMotion: Motion = Motion.coin(

1.0 / 33,
 Motion.linexp(Motion.walk(0, 1, 0.1), 0, 1, 0.4, 4.0), // Motion.exprand( 0.4, 4.0 )
 Motion.exprand(4.0, 16.0)

// val matchDeviationMotion = Motion.linrand( 0.2, 0.5 )
val spectralMotion        : Motion = Motion.linrand(0.25, 0.75)

val stretchDeviationMotion: Motion = Motion.walk(0.111, 0.333, 0.05) // Motion.linrand( 0.2, 0.5 )
val rankMotion            : Motion = Motion.linrand(0, 11)

val maxBoostMotion        : Motion = Motion.constant(30) // 18
val minSpacingMotion      : Motion = Motion.constant(0.0) // 0.5

val minPhraseDur = 10.0

def findMatch(rank: Int, phrase: Phrase[S], punchIn: Span, punchOut: Span,

              minPunch: Long, maxPunch: Long, weight: Double)(implicit tx: S#Tx): Future[Match[S]] = {

  import synth._

  val maxBoost  = maxBoostMotion.step.dbamp.toFloat
val minSpc    = secondsToFrames(minSpacingMotion.step)
val dirFut    = db.asStrugatziDatabase
val metaFut   = phrase.asStrugatzkiInput

  dirFut.flatMap{ dir =>
   metaFut.flatMap { metaInput =>
     findMatchIn(dir, metaInput, maxBoost, minSpc, rank, phrase, punchIn, punchOut,
       minPunch, maxPunch, weight)

Code fragment 2011 (source)

like any memory, the view of an old piece mutates over time; one can try to remember how something felt back then, but actually one feels different. how I felt about reading the Grammatology back then, and how I feel about it (no longer that interesting). but that is also the charm, that you would not do it again in the same way, if you had begun to work on it today.

you could not do it again in the same way.

the gap between reconfigurations also indicates the algorithmic as a medium; the potentiality of rewriting and discrete shifting. changes happen both in the discrete (separated) and under discretion (tacit). The discretion is an ambiguity that is only seemingly resolved through statement, because the statement pretends to resolve something that in fact remains unsaid. The observation is not innocent. When I correct myself, the observation is responsible for the manifestation of the change. This is one side of the change, in which we are passive or reactive. We make a move, but this decision is consequence to an alien agency or will that encircles what is possible.


{keywords: [memory, past, presence, time, reconfiguration, algorithmic, medium, rewriting, writing, observation, agency]}

the singular “writing” is always suspicious, as the source of truth and authority, just as the most widely spread religions are based on monisms. the multiple on the other hand(s) seems like the supreme expression of the algorithmic. as a potentially endlessly ongoing writing, iteration, recursion, reinstantiation. as Frieder Nake put it: “The work of art in algorithmic art is the description of an infinity of possible works.” perhaps against the backdrop of capitalist-industrialist automated production and techno-optimism, when Cage was interviewed about HPSCHD, he said that “in the case of working with another person and with computer facilities, the need to work as though decisions were scarce—as though you had to limit yourself to one idea—is no longer pressing. It's a change from the influences of scarcity or economy to the influences of abundance and—I'd be willing to say—waste.” waste, surplus, excess. it's interesting that Cage here likens the work on computers with collaborative work. algorithmic aesthetics is still full of suspicious monisms, as one can readily see when reading Nake. there is still the idea that artists' genius formulates the one description of the process. his heroes Turing and Church “saved mathematics as the only discipline of the human mind that can say clearly what it says.” in contrast I would say, the algorithmic is inherently about the multiple and escaping any attempt to return to the singular “algorithm”. the multiple is in the trace, while the origin never existed (Derrida), the trace performs a play of difference and repetition (Deleuze) or differential reproduction (Rheinberger).

an interesting new possibility is added through simultaneity. no longer iterating step by step, but assembling things to happen at the same time, together-in-independence.


{persons: [Frieder Nake, John Cage], keywords: [writing, monism, algorithmic, iteration, recursion, description, possibilities, artwork, computer, collaborative, aesthetics, process, multiple, multiplicity, simultaneity, independence, iteration]}

it could be seen as an inflationary use of the term. if it's broken down to inscription (as in N Katherine Hayles' ‘Writing Machines’), or imprint, then of course any form has been generated by some sort of inscription.

how do you perceive a form? by de-scription (Be-schreibung).

I would argue that all the activities, such as understanding the materiality of the copper wires, how you can bend and solder them, how they arrange “themselves” as part of their stiffness and elasticity, how they reflect the light, how they create a transparent layer, etc. that these “readings” lead to a particular form of “writing” (and now I need to be careful not to say “language”) in the display of the piece.

a writing does not always become apparent to others, at least in the first encounter, but somehow writing as a ritual becomes important for me to perform when developing a piece. for example, finding the wood blocks on which the relays are displayed, finding the blueish gray colour for the wood blocks, choosing the type of coarse black nails to affix the relays; this may be a minor visual detail that one would probably overlook, but these small “auxiliary” forms are important to me. the heap of electronics below the glass table is a sort of landscape.

we should not think of writing as something asking for deciphering.

the previous iteration had the piezo discs soldered in groups of six (I believe), and the number discs was lower (eighty-something) than now with three tables à 27 discs, I didn't have 91 of these piezo discs. So I needed to order more discs. The existing piezos were bought in 2011 or even earlier, and I could not remember where I bought them, nor did I find this model in the common electronics suppliers. I could come as close as finding discs of the same diameter, but without the white plastic ring around them. It turned out that these new discs refused the solder that I had, it was impossible to attach the wires. But I had already started to remove the white plastic discs to make them look the same, so I completed that step. Luckily I had another piece using the old discs that I could cannibalise now. I never liked the white plastic. The person that was helping me with the construction work did not understand why I wanted to scrape off the rubber that was used to attach the plastic discs. Because the effect would be invisible, as the discs would eventually be immersed in the graphite heaps. However, I found it interesting to see the difference between the plain metal and the outer parts that were scuffed from the cutting knife. I was already thinking of a future use that would show those scratches. While setting up the tables, I was lying on the floor a lot, fiddling with the electronics and the cables. One can look upwards from that position and see the entire table arrangement from the bottom. I thought that was a beautiful view, even if few people would see it this way. I was happy that I also enforced a particular way of glueing small plastic segments on the discs so they could be placed flat with space for the wires and the solder points. Now it was important. It is always worthwhile to insist that you pay attention to the “auxiliary” forms.


{persons: N. Katherine Hayles, keywords: [writing, machine, inscription, imprint, form, description, materiality, copper, reading, ritual, performance, electronics, deciphering, piezo, plastic, rubber, metal]}

meta: true
event: almat2020
kind: essay
date: 200919
author: HHR
place: Online

keywords: [writing, artistic research, sound installation, algorithm, machine, rewriting, artist, simultaneity, making, presenting, research, traces, material, computation, computer, space, togetherness, co-writing]

def selectPart(dbFile: File, phFile: File, instr: OverwriteInstruction): GE = {
 import de.sciss.fscape.graph._
 val dbSpec = AudioFile.readSpec(dbFile)
 val phSpec = AudioFile.readSpec(phFile)

  require(dbSpec.numChannels == 1)
  require(phSpec.numChannels == 2) // left channel is sound signal, right channel is 'withering'

  def mkDbIn() = AudioFileIn(dbFile, numChannels = dbSpec.numChannels)

 def mkPhIn() = {
   val in = AudioFileIn(phFile, numChannels = phSpec.numChannels)
// there is a bloody bug in fscape audio-file-in with the second channel dangling.
// this seems to fix it
    Length(in).poll(0, "length-ph-in")

   in out 0

// // XXX TODO --- enabling this prevents some hanging. but why?
// // if (Main.showLog) {
// in.poll(0, "ovr-fsc")
// // }

  val stepSize    = fftSize / stepDiv
  val sideFrames  = (SR * sideDur ).toInt

 val sideLen     = max(1, sideFrames / stepSize)
val covSize     = numCoef * sideLen

// - punch-in : start = max(0, min(phrase-length, instr.span.start) - sideFrames); stop = min(phrase-length, start + sideFrames)
// - punch-out: start = max(0, min(phrase-length, instr.span.stop )); stop = min(phrase-length, start + sideFrames)
// - from this we calculate the offset between the two taps into the db: delay-length = punch-out.start - punch-in.start
// - from that we calculate the run length: db.length - sideFrames - delay-length; if that's <= 0, abort here
// - from that we calculate the number of window steps (/repetitions)

  val phraseLen = phSpec.numFrames

Code fragment 2019 (source)

Three separately recorded tables

simultan 1


simultan 2


Mellite workspace with code fragment 2020 (source)

Experiments based on the sound piece “Unforeseen - Real - Achromatic”