Sounding the Net: Interview with Jon Thomson

Peter Traub: There is something very Cageian about the piece, obviously in your reference to prepared piano, but also in the chance nature of the work. Then there is also this notion of this very traditional acoustic instrument connected to and playing back the sonic ephemera of this technologically sophisticated global network. Could you address both of these observations, perhaps discussing why you would or would not agree with them, and how you thought about the piano as a 'voice' for these net-based ephemera?

Jon Thomson: As you suggest, the Cage reference is implicit. The title is both a joke (the idea of the piano being unprepared for what it is instructed to play) and a conscious connection to the idea of prepared pianos --it's just that in this instance, the preparation is by means of software --a set of rules that characterise the way the piano adapts any rendition of any given 'found' midi file. Again, you are right that we chose a midi grand piano precisely because we wanted to pit the generative playback against the authority such a traditional acoustic instrument maintains. We wanted this authority to goad listeners to 'read' these musical works within that context. However, we don't really think of the piano as a 'voice' for net-based ephemera -the web in this instance is a data base or library of scores that the computer picks up on and draws from. We like the openness of this connection, and that the midi files are 'naturally situated' online, but in our minds it's not the most important aspect that it is the net, but rather that the piano is connected to a very large body of data, from which it selects and performs. A large enough body of information to make the permutations of rendition feel almost endless..
    PT: This is interesting because it speaks to a question I had for another interviewee. I was wondering about networks as a subject in pieces vs. networks as just the medium or data source in which the network itself isn't really central to the idea of the work. From your answer above, it sounds as if the network is really just a tool or an access point to really large data repository to feed to your piano. Have you noticed a shift in your internet or network-based pieces in a way that you are perhaps less interested in the 'network' aspect of the piece now than as using the network as a broadcast medium or data source, etc. i.e., the piece doesn't need to be about the network or the nature of networks, but yet the network is still central to the piece's existence? I am particularly curious to know how your thinking about creating online works has changed over the past 5 years or past 10 years.

    JT: There has been some shift in our work in recent years, where we are looking more and more at possibilities for networked installations that might appear in a gallery for example. Decorative Newsfeeds, Short Films about Flying and Unprepared Piano all fall into this category -and yes in the case of the piano, the net is a convenient repository of data, and although instrumental (no pun intended), it is not using the web as a specific site. This hasn't been at the expense of our interest in making art works that inhabit the web, so should be seen as an expansion of interest we guess :)
PT: Just to jump back a bit and ask the obvious, how did you come up with the work and what sort of compositional approaches did you take in creating it? Being that it deals with a lot of chance material, what sort of compositional decisions did you make or do you feel where most important in creating the piece. Feel free to refer me to a previous interview or article if you've already discussed this at length elsewhere.

JT: A computer is connected to the web. It finds midi files that are multi-timbral and downloads them adding them to a data-base. A max patch then loads a file, and plays it from beginning to end by performing a random number of channels for a random amount of time within set limits (a min and maximum amount of time). These limits change at random times during the course of playback, which also fall within set limits (minimum and maximum), while the speed the piano plays the file at, also accelerates and decelerates according to the same kind of randomness set within minimum and maximum parameters. The result is that the piano plays the whole file from beginning to end in snippits and combinations of instrumentation -bits of marimba, bits of bass, bits of piano, bits of strings, bits of drums &c, so that while the resulting 'music' is quite distinct from how the original score is likely to sound, certain dynamics and traces of the originals structure remain. This we think of in a similar way to the inherent authority of the piano --as innate qualities in a way. These rules are our equivalent of Cage's physical preparation of pianos, because it is a consistent template, (even if within that it operates with randomness). This is a similar idea that drives our interest in making templatecinema.com and an installation called, 'Short Films about Flying'
    PT: Yes, I watched one of the 'Short Films about Flying'. I'm wondering how you think about serendipity and/or 'happy accidents' in your work. In chatting with Jason Freeman, who wrote N.A.G. (Network Auralization for Gnutella), he said he was particularly interested in the serendipitous results of searches that return unexpected data (i.e., a search for 'madonna' on Gnutella may bring up any song by Madonna but also the Beatles's 'Lady Madonna', etc... and those would be mixed in the application). There has been a very similar theme in my net sampling works, and I'm wondering if you two also think a lot about serendipitous events? 'Short Films' seems to imply that you do, in a similar way that Cage perhaps thought about them in pieces like 'Radio Music' and 'Imaginary Landscape No. IV'. Do you think that the semi-chaotic nature of networks somehow lends itself to works centered around serendipitous events?

    JT: well in general I think you would be hard pushed to find any artist who doesn't at least respect the concept of a happy accident. In the case of Short Films about Flying though, what we were particularly interested in was the role the observer/viewer plays in the formulation of the status of these movies i.e. whether they mean anything or not. Although each film generated follows exactly the same structure (2m 45s long, four inter-titles inserted at fixed points, sound track, titles and credits) the elements within that structure are called up randomly making every movie unique. Yet the authority of the cinematic format and our desire to make meaning out of everything seems to will meaning into them, almost. In a way it is the viewer that ends up tying these disparate elements together in their natural desire to make sense of things.
PT: I'm unfortunately unable to access your site right now (browser just hangs while loading it),

JT: fixed now. our server was playing up at the weekend.

PT: as I wanted to watch the movie of Unprepared Piano again before writing this question. So please forgive me if my memory of what I saw is a little off. In the movie, the piano appears to be set up in a very public space, like a train station or airport terminal. There seems to be something very anonymous about it, much like most of the material that it is gathering. Can you speak to the idea of this work existing in a very public space and its relationship to that space and to the public network from which it draws its material? How do notions of anonymity figure into this, if at all?

JT: The space was a huge hall in Paris called Grande Halle Parc de La Villette, and was part of a big exhibition there. We do like the idea of it being in a restaurant or public space, and just babbling to itself, but is not really part of the intent we have for the work. Anonymity is not something we've thought about a lot in relation to this work, but we will do and maybe comment further down the line...

PT: The network is your source of material for this piece, and I am wondering how you thought about the network in creating the piece. It isn't really the medium in the way that it is for a piece of net art, but yet it is vital for the existence of the work. In the piece, the listener just hears these seemingly random piano gestures and pieces coming forth without any knowledge of where they came from (or is that presented on a monitor nearby?). It seems with that kind of arrangement that you could use any number of methods for generating material other than the internet (a Max patch for example), and the end result would be similar. I bring this up because I'm working on a piece right now in which the identical argument could be made. So my question is, why is the network relevant here?

JT: I agree that the infinite mutability of data is both promising and problematic. With Unprepared Piano, the piano stool is absurdly kicked aside and a monitor display occupies that space, and displays status information about what's going on: the name of the file, the web address it was found at, the total number of events in the file so far and the total number of events the piano has played so far. This is hopefully enough to ground the viewer/listener and give them some sort of explanation as to what is going on. Obviously you have to just accept what's being presented on screen as what's actually happening but that need for trust is true of many artworks --especially conceptual ones. The status screen also forms a kind of artificial subtext. If the file is called salsa and it comes from a website called angelfire.com/peteslatinexpress/salsa.mid then it affects the way you read it/listen to it.
    PT: Is the installation constantly searching for new midi files? I suppose this goes back to the network-relevance question, because there is clearly this importance of having the installation searching for and pulling these files in realtime as opposed having a database filled with thousands of files that were pulled from the web months ago. Is the realtime aspect important to you, in having this machine connected to a live network pulling files from it for source material as it finds them? What I'm trying to get at here is 'why be connected to a network?'. If the piece could run off a database of previously found files (but it doesn't), then there is something important about that live connection to you and to the piece. Could you speak to that a little bit?

    JT: This is probably not very helpful, but it can do either as things stand. To make matters worse it can also pull from grouped sources located online, so we could have our own library on-line and get it to choose from others too but according to how we configure the patch. As a discreet work I suppose it doesn't really matter too much to us whether Unprepared Piano is connected or not -it is important however that the information describing where a file was sourced originally is maintained for the sake of the status display. So perhaps there is your answer.

    However, as and when we combine it with live web cam images and text, to make a new template cinema work with the piano as soundtracking device, it will be important that it all grabs live and in realtime. It is at this moment (as with templatecinema.com) that we become really engaged with the idea of making movies that are simply a list of instructions and nothing more. We liken them to Sol Lewitt's instructional drawings, but in this case of course our instructions execute themselves. Another metaphor we use is that our template cinema movies are like knots that tie unrelated threads of data together for a moment and then allow them to fall back into disarray again. We're about to publish a book chapter where we talk to a sociologist about this. It's forthcoming but I could send you a transcript if you want, as long as you cite it properly if you use quotes (i can send details of that on request)
JT: (addendum) One other general thing about Unprepared Piano. We originally thought of making it because we wanted to create a different kind of soundtracking device for our ongoing template cinema work. In a way, the piece is a prototype for later work that uses the piano alongside live real-time visuals being pulled from the web. As and when we manage to make that work, Unprepared Piano (as a component of a larger template cinema work) will be much more intrinsically linked to the idea of the web.
    PT: Ok, here are questions regarding "Telephony"

    PT: Were all the phones in the grid accessible to outside callers, or did you just want a few phones as entry points? When a single caller dialed in, how many times would that call be forwarded to other phones on the grid before a cycle ended?

    JT: All the phones are live, so anyone can call them. You get quite a few wrong numbers too that come into the space -in that way the grid is like a net that catches them. Also call registers can create odd effects. For example, if someone dials out from the grid by pressing the call button twice, it calls the last phone that called it. So we had instances where people who had used the work at the opening let's say by dialing in on their own phone, would keep being called by the installation! Worse than that even, the second time we staged the work (some months following the first showing) we forgot to delete the call registers, so people who had visited it first time around were called again.

    As for call forwarding, it depends on the network, but on average it will call forward 5 - 7 times then give up the ghost.

      PT: I find this a very interesting phenomenon. Golan Levin described a similar phenomenon with his work 'Dialtones', whereby people who had registered their phones to be in the audience ended up not attending the performance, but during the performance these people were called by the performance software, which assumed they were part of the matrix of phones in the performance space. It somehow really illustrates this connection (or perhaps disparity?) between real and virtual space, such that the participants had left the real space, but couldn't leave the virtual. In a similar way, this seems to be the case for those people who were called back months later from 'Telephony'.

      JT: Yes exactly. We often talk about the the way in which mobile communications technologies extend an individual's sphere of influence (the converse being true also), where you can be physically somewhere but virtually anywhere. This also taps into ideas of translocal &c.

      PT: You appear to describe this effect in a negative light, and perhaps it is such for those who are called back at midnight from a gallery installation, but I would think this might almost be considered a feature rather than a bug. In my opinion, it sounds like quite a wonderful trace or ghost effect of the piece. How do you and Alison think about it (it seems to run into my earlier question about happy accidents)?

      JT: No it's not negative. Although unanticipated it is exactly the kind of effect we want to discover when using the telephone network as a site for an artwork.

    PT: Being that the Nokiatune is such a recognizeable commercial jingle, it seems to offer certain challenges and benefits in appropriating it for a piece like 'Telephony'. In creating variations on the Nokiatune for your ring tones, how did you think about the source material compositionally? What types of mutations/transformations did you apply to the tune, and was the process algorithmic in any way?

    JT: It wasn't algorithmic. We just composed them ourselves trying to take certain things into account -ie keeping in key, keeping the variations related to the structure of the original, and making sure it related well to the background musac. Because the order and timing and layering was beyond our control we kept them simple and restricted ourselves to fairly brief loops. However, in telephony, the music is not what's important, it's about situating phones in a gallery and scrutinising the way we behave towards them, and how the live network is instrumental to how the work operates, and how people experience it. It's a kind of invisible drawing in our minds.

    PT: Is there any sort of underlying network structure to the phone grid that would not be obvious to a viewer/listener? i.e., a phone always forwards to the phone one over and one down from itself or something to that effect. Did you design some sort of connection/relational pattern between the phones, and if so, were the ring tone variations tied into that pattern?

    JT: We've shown it a few times now, and when we first did, the connections between phones ran from top left to bottom right. This was quite nice, because when a phone is called it lights up, allowing for an interesting animation when more than one person dialed in at the same time and as they call forward. The screens become a bit like pixels or a kind of score after the fact. However, more recently we've just mounted them randomly, which also works for us. Not many people pick up on this visual aspect -sometimes people are just interested in calling / texting their friends from the wall. This is something else that is important to us, namely that the system is open. You can use the phones normally, well as normally as you can use a phone glue gunned to a wall in a grid of 42 :) We really are interested in seeing this as a work that hangs within the live network -back to the drawing metaphor again.

      PT: Yes, the drawing metaphor. You mentioned it in a previous answer, but I wasn't quite sure how to respond, but perhaps you could clarify this a little more. In terms of the drawing metaphor, do you see the intangible structure of the piece (the way the network works, people respond, etc.) as sort of living drawing, or are you talking in more concrete terms, i.e., the grid of 42 phones on a wall is like an active/networked painting in a gallery and should be viewed as such?

      JT: I suppose we are thinking more about the former. We enjoy the invisible material presence of the nodal points that can be identified within the networked structure of a piece like telephony and how that implies the linear connections between the points as a drawing or diagram -- harking back to or making links to historical references like, 'from A to B and back again' or, 'taking a line for a walk'. In the case of our work, 'Decorative Newsfeeds' we often call it an automatic drawing, and of course as mentioned already, we reference Sol Lewitt's instructional drawings with our template cinema work. More than aphilisophical stance or consistent interest, we think that our use of drawing metaphors are mostly due to our background: we both went to art school and started with drawing and painting before moving into video then use of communications techonologies, so old habits die hard :)

    PT: An article I read about you online mentioned part of your interest in the internet being that it is not perfect and that is filled with outdated browsers and computers and web pages, etc. (I hope this characterization is correct, please let me know if its not).

    JT: well that's true to a point -where did you read this?

    PT: This seems to tie in with another interviewee who was particularly interested in the fallibility of networks and the internet. The lie detector you use in 'Telephony' also somehow seems tied into this notion of machines and networks being sources of data but yet very fallible sources that must be carefully scrutinized (almost human in their fallibility perhaps?). I'm not sure if I'm making sense with these connections, but I hope you see where I am trying to go. I am wondering how these ideas manifest themselves in 'Telephony' through the use of the clocks, the phone bank, and the lie detector?

    JT: In a way, the lie detector testing speaking clocks, is a separate work to the telephones, even though we present them together more often that not. When they are together, the voice of the universal clock (currently not available online anymore) becomes like a component of the 'music' being produced, where the supermarket musac, ringtones and real audio stream all act together to make one audio work. The reports that the lie detector generate however are exactly about revealing the fallibility of networks and data sources as you put it. By pitting two didactic devices against each other (voice stress analysis and the speaking clock) both end up undermining each other. For example, the British speaking clock is deemed, 'prone to exaggeration' while the Irish speaking clock is, 'unsure of much of what it had to say'

      PT: It is interesting what we consider fallibility though. With respect to the lie detector, can you really say it isn't working properly, or that it just wasn't given data that it could accurately analyze given its ruleset. To some degree, this type of fallibility would almost be predictable in that you could determine what types of voices and data would make it give false positives, etc.

      JT: Indeed, but it reminds us of all the false positives given but obscured when the software is used in its intended way on a human subject. The software we use has been considered legally legitimate means in some countries and sentences have rested on the results.

      PT:This seems to also be the case with the ghost phone calls you discussed above. Was the network/piece really failing in this sense, or was it actually doing exactly what it was told to do, but just in a way that was unforseen by its creators?

      JT: that's it.

      PT:I suppose I am reluctant to call those things failures, but I'm wondering if you'd disagree? Is it a different type of failure from a poor network connection in which packets are lost and data transmissions are left incomplete?

      JT: This is a bit pedantic because we pretty much agree with your stance, but technically we would disagree with you as we are less reluctant to use the term. The word, 'failure' is quite a useful one in our minds when attributed to the, 'functioning' of an artwork, because we think a great artwork can 'fail' just as a poor one can, 'succeed'. A generally loathed landscape painting might succeed in representing an image of, or reflection of a place by being depicted in such a way that most people recognise it as being 'of something'. While Douglas Hoebler (for example) fails to take photographs of absolutely everyone in the world, while still making a coherent series of artworks..