Friday, 29 March 2013

Something I wrote recently. There is a short film it accompanies, which should follow soon...

Cleansing: The Silicon Roundabout
The Clash of Myth, Gentrification, and Community 

The drive to ascribe a near mythical status to the cluster of new, creative companies in East London has grown to a cacophony There is an explicit political and economic motive to have the area around Shoreditch viewed as the place to be for companies perceived to be brave, trendy, and cutting edge, and to some degree it is working. Whether the chicken came before the egg is debateable, but there is now a concentration of both indigenous start-ups and vast multi-nationals like Google and Facebook trading on the capital of the Silicon Roundabout.

Shoreditch, which accidentally found itself to be the heart of London’s creative and artistic scene in the 90’s, was until recently a firmly working class community, and its industrial past is still present in the converted loft apartments, open plan ‘studio’ offices, and the ‘industrial chic’ of so many of it’s restaurants and bars. These re-invented spaces reflect the huge demographic evolution of the area, which has caused many to stop and question the narrative of inevitable improvement that accompanies these waves of gentrification in urban areas. 

One such person is Robin Bale, an artist, poet, performer and local resident who has taken particular interest in developments. His first work to directly reference the changes occurring was titled ‘A Circumvention of the Perimeter of Hackney Council’s Designation Orders’ (2009), in which he led a walking tour/performance of the soon-to be no drinking zones in “Designated Public Places”, reflecting what he perceives to be the institutional social cleansing of the indigenous working class population for the benefit of the new ‘consumers of surplus value’ that now inhabit the area. Bale, with his appearance mirroring that of a street drinker himself, is trapped between the two worlds, and doesn’t appear to fit naturally in either. He could be taken as a metaphor for Hackney itself as it grapples with the opposing forces of community, authenticity, creativity and cold, hard economics. A recent theme in Bale’s work is the relationship between data and financial flows; what he perceives to be the only two remaining forces in the area. 

The dominant current of cultural theory resolves around the conception of networks as the defining social relationship of our time, and all the rich metaphors they produce. Through this lens, we must end our treatment of cities as places, and begin to understand them as process, one in which people, spaces, and cities define each other by their relationship to one another. This ‘space of flows’ as Castells sees it, is defined by layers of material support for the accompanying social practice: the base layer is grounded in materiality, and constitutes the infrastructural support for all else from cables and wires to jet planes. The top layer is constituted by our new techno-financial-managerial elite, and is manifest in the homogenised spaces they occupy. The middle layer is the glue that brings it all together. It is the space of social practices that define society, and the link between physical spaces that enable economic, political and cultural functions.  By this rationale, areas with high concentrations of networked, successful companies create a momentum and a force of their own, becoming key nodes in “The Network”. There is an argument that the reality of the cluster isn’t even important; the idea of a cluster can be enough. True or not, what is clear is the political drive to brand this particular part of east London as one such node. And the residents of the node need space, space that is currently occupied by thousands of local residents living in one of the poorest areas of the UK. Since the birth of the modern city, the middle class need for ever more space has been abated by moving into previously undesirable areas.

It is estimated that as much as 40% of Hackney’s working class population has left the area in the last 15 years for reasons including evictions, demolitions, rent increases, and the transfer of vast swathes of council housing into the hands of housing associations. This is part of an openly pursued policy of gentrification by the local council. The borough is scarred with derelict and almost empty housing estates, many of which only hold off demolition due to the occasional stubborn leaseholder or – the irony is not lost – the credit flow of developers in these straightened times.  The developments that replace them are contractually bound to provide a percentage of ‘affordable’ housing in return for the planning permission and tax breaks they receive for their apparently benevolent role in the community. However, by the governments own – if unloved – research, the classification of ‘affordable housing’ refers simply to a specific type of home, rather than one that is actually affordable in any kind of economic sense or understanding. Of the half dozen housing developments underway, that average about thirty stories, in the immediate vicinity of Old Street, none are expected to have homes available at anything less than £200,000. Thus regeneration has in fact served to accommodate the theft of the city from its own occupants, re-engineering areas for young middle class consumers and their Macbook pros. This is particularly manifest in the no-drinking zones, and it is clear why Bale took quite such exception to them. 

It comes down fundamental ideas about how we understand our relationship to the city. Public spaces are just that: public. Streets, parks, squares and roundabouts are spaces to be used simply by whoever happens to be in them. In Hackney’s no-drinking zones, the rules applied in ‘Designated Public Places’. Who is in charge of the designating? We then find ourselves having a ‘Designated Public’; those members of society that behave in a manner constituent with the bylaws in those spaces. Clearly it will not be the occupants of the new developments (on top of the old estates) that break the rules; the rules are designed to differentiate acceptable and non-acceptable ‘publics’. All this arrives at a time when choice is the overriding mantra of the day, when it seams the creation of choice is all a leader needs to get elected. But we can’t choose to drink a can of lager in the park if we so fancy. And the opposite of choice is dependency, the discursive validation of the social cleansing we see. Dependency reaffirms our notions of worthy and unworthy poor. These are all themes that rise and fall like the tide in Bale’s improvised performances, much to the chagrin of the trendy young things around him on the streets and back alleys of Shoreditch; he gets quite animated. Can’t anyone just get a Frappuccino in peace around here anymore?

The myth that is the Silicon Roundabout is about to produce a brutal glass and steel metaphor of itself. Recently, David Cameron and Boris Johnson gleefully announced funding for a £50m ‘renovation’ project for Old Street roundabout, the concrete underpass, shopping centre and tube station on the junction between London’s financial district and the Borough of Hackney which has, by default, come to represent the concept of an East London ‘tech hub’ in the minds of those lacking much semblance of imagination. The council itself describes the roundabout as “disparate in form, scale, age and materials with very little that is memorable”, but all that is set to change. This edifice will take the place of the misguided steel sculpture that currently occupies the space, and is most often taken for an elaborate billboard frame; recent history’s attempt at ‘Public’ art. This new ‘civic’ space will contain areas as fantastically useful as a  ‘Sandbox’, a ‘Fab Lab’ and a ‘Think Garden’, amongst others. Unfathomable names for intangible assets. When civic space is so explicitly exclusionary how are we to make sense of the community it is said to be provided for?

Is it possible to create ‘community’ if we wish it hard enough? We had better hope so, as the wholesale destruction of existing communities has already begun in order accommodate the new. The diagnosis of the city as existing in a web of nodes and networks holds its strength, but we cannot live entirely in abstracts. Indeed, we can agree with Benedict Anderson that communities are discursive constructs, but that does not make them any less real, and it does not negate our responsibilities to those that happen to reside where we desire to construct the next one. 

Click this and listen whilst you read, you will not be disappointed....

Sunday, 17 March 2013

Haiku Whisper’ – an (inter)active audio installation

Poetry as performance is a diminishing art form, particularly in a society that is moving further into the separate physical spaces of individual media devices and away from shared art experience. It is suggested that this is particular manifestation of a wider disconnect with our auditory environment. Into this context is presented Haiku Whisper, a poetry generating system designed to create unique poetry in the Haiku tradition to an unsuspecting audience. Furthermore, the poetry is created in response to participant’s movement through physical space, and played back in real time; a pause in movement creates a pause in performance. It is hoped that such a system can reintroduce an overlooked art form and induce a heightened awareness of our personal soundscapes in a playful, creative way. This paper considers the academic imperative for the development of such a system, presents the technical aspects required for execution, and discusses the reception of the preliminary public installation. 

Wall - sensor - Arduino - MaxMSP - speakers - wall. It's as simple as that.

Haiku - (ai haikai) is an ancient Japanese form of poetry, consisting of 17 on (also known as morai), which in English translates – almost - as 17 syllables, put together in three phrases of 5, 7 and 5. The essence of Haiku is in the juxtaposition of two images or ideas, and a kireji or ‘cutting word’ between them. The fundamental aesthetic quality of Haiku is that it is seen as internally sufficient and, despite its brevity, can bear consideration as a complete work. Furthermore, each of the three lines is considered to represent a conceptually complete thought or action. These two features contain potential for creative manipulation whilst maintaining the integrity of the tradition. 

Haiku Whisper an interactive oral installation - with orality understood as both the act of speaking and listening [5] that generates and performs random yet structured Haiku poems to unsuspecting audience members as they move through a public space. It is an attempt to re-introduce an unsuspecting audience to a cultural aesthetic that has been losing ground in contemporary society, and by choosing Haiku as a platform seeks to reflect an increasing cultural interdiscursivity. Furthermore, it attempts to do so whilst posing the audience questions about how the understand their everyday auditory landscape.

Poetry as communication has existed throughout history but, has lost traction in recent times; as a form of entertainment is has become increasingly confined to taught environments and literary circles [8]. Furthermore, as a society we are becoming increasingly removed from our immediate environments as we withdraw into our digital devices, becoming ‘alone together’ [9]. 
With these twin premises as the point of departure, the interactive installation Haiku Whisper was designed to investigate the latter with the use of the former. By using hybrid aesthetics that reflect the character of the modern generation, the hope is to make the audience question their assumptions of their everyday auditory environment. The primary intentions of the project are:
  • To communicate a cultural aesthetic that has been losing ground.
  • To create an interactive system that faithfully generates unique Haiku in a coherent, sensical form.
  • To create a sense of ‘belonging’ to the interaction; the user should be aware of their role in the piece, even if their understanding does not amount to control.
A strong and varied body of interactive poetry generation systems has appeared in the recent past, with varied success. Tosa [7] developed the Interactive Poetry System in which human and computer agents collaborate through the exchange of poetic phrases, whereas Rokeby’s The Giver of Names [1] generates poems based on objects selected by participants and collated on a pedestal. Zhu et al [10] created Poetry Mix-Up, in which users create a ‘re-mixed’ form of poetry by communicating with the system via Short Messaging Service (SMS), bringing a social communication value to the interaction and grounding it in modern culture.  With Hitch Haiku [8] Tosa et al present system that generates Haiku based on participant selected phrases from a famous Japanese essay (1000 Books and 1000 Nights), translating the outcome into English. Although Haiku Whisper is a less complex – and therefore arguably less nuanced – generation system, it’s strength and advantage lies in the ability to connect the participant to their physical and audio space in a manner that has not been addressed before.

System design
Oliveira defines four categories of poetry generation systems, defining Haiku Whisper as ‘template based’, in which systems fill templates that carry syntactic and/or rhythmic constraints [4]. Template based systems are acknowledged as the least complex, but suit the design requirements adequately. A more nuanced approach may be suitable for later elaboration.

A series of imbedded Passive Infra-Red (PIR) sensors register movement through and Arduino microcontroller. This in turn triggers the random selection of audio files contained within a Max/MSP patch, feeding back through a multi-channel audio system to a series of paired speakers matched with their corresponding sensors. The PIRs and speakers are embedded within the external walls of a busy London street. 

The Max Patch

There are three PIR/speaker units, accounting for the three lines constituting a Haiku. Distances between three and four meters separate these. When triggered, the corresponding sound file plays, and lasts between two to three seconds. This correlates with the time it takes a fairly fast paced pedestrian to reach the next PIR, thus triggering the following Haiku stanza in time and rhythm with the last. Therefore if the participant walks continuously, they will hear they will experience their uniquely generated haiku spoken to them as they move through the space. In contrast, if they are to stop or pause, so will the performance, logically only resuming as they do. 

The sensor
The sensors are HC SR501 PIRs, selected for their high sensitivity, diminutive size, and low cost. Furthermore, they are equipped with two variable resistors to calibrate both re-trigger delay and range and are also highly compatible with the Arduino microcontroller. The HC SR501 has two IR detecting lenses; when a body passes it first intercepts one lens, causing a positive differential range between the two halves, thus triggering the microcontroller. The sensors were placed in custom built cases and mounted in 200mm steel tubes to reduce the sensitivity from a 110° field to a band of 50cm at a range of 4m.

The database
Haiku is a language of abbreviation; it is an art form that infers a wider story from the context it creates, allowing for great leaps of imagination and disparate thought processes. As such, a database was created that would reflect and embrace the inter-changeability and idiosyncratic possibilities of the form. The preliminary database used 160 original recordings of genuine, published Haiku mixing both translations from Japanese and those composed in English. Earlier tests experimented with online digital generators, but the results were of a low quality and as such rejected. These original recordings, performed by a professional actor, proved invaluable in ensuring clarity and quality of reproduction, particularly in consideration of the high levels of background noise encountered when the installation was active. Furthermore, Manurung [3] affirms that a poetic text must hold the three properties of meaningfulness, grammaticality and ‘poeticness’; the database is selected to reflect this. 

System evaluation
The technical aspects of the system work well, with a 90% successful trigger rate and no conflict between triggers/sound files being activated concurrently. The only failing with the system at this time is that it only works fully when the participant walks in one direction (in this case East to West) i.e.: files are in grammatical order - 1st, 2nd, 3rd – when triggered from Left, Middle, Right. The system still produces rhythmically complete Haiku when triggered from either direction, A change to the software could reconfigure this, so that the stanzas are triggered in order of poetic occurrence rather that which PIR sensor they are assigned to, but this brings forth new issues concerning both re-setting and multiple participants in the space at the same time. It is therefore proposed that a unidirectional installation is preferable to such a potentially chaotic one. 

In terms of actual poetic output, the system produces complete, yet not always coherent, Haiku poems. There is an obvious element of abstraction and surrealism to the output, but this was fully expected and even encouraged in the database selection. In addition, due to the extreme brevity inherent in Haiku, it is by definition abstract and relies heavily on symbolism for it’s impact; when there are so few words to available, what is inferred is often more important than what is said [2]. In addition, the structure of Haiku lends itself to the type of manipulation undergone in Haiku Whisper. As stated above, traditional Haiku entails the juxtaposition of two contrasting images or ideas – contained in the first and last phrases – and a kireji or ‘cutting word’ between them, which defines the manner in which the juxtaposed images are related and interpreted. This is generally manifest in the middle phrase. Therefore, it is perfectly feasible to switch phrases between Haiku whilst maintaining the poetic integrity and tradition of the form. As such, Haiku Whisper should be viewed as a success in terms of producing output that can confidently be identified as Haiku. 

Observational evaluation
The installation was a site-specific design for placement on Mile End Road, East London, E1 4NS. As such, the installation was activated and observed for 2hrs on Sunday 17th February 2012 from 1-3pm. During that time, 32 people became participants in the piece; 17 single, six pairs, and one trio. Their physical reactions were observed and recorded.
Of the 32 participants, all were observed to react in some form or other. The smallest reactions involved simply turning the head in surprise or expectation of finding someone close by who had spoken the words themselves. Other reactions included a slowed or quickened pace, stopping entirely, inspecting the wall in which the audio units were embedded, and calling out to question who was there. An initial shock was observed on several occasions, as participants were seen to jump or turn sharply when the first stanza played. 

The majority of participants understood that the audio was coming from the wall of the building, and those that were in pairs or more we often heard to be discussing as much. Of the 32, those in pairs (or more) were seen to be more inquisitive as to the source of the noise (although this is undoubtedly due to other social factors and not the inquisitiveness or lack thereof of individuals), and three pairs and the trio investigated long enough to discover that it was in fact their movements that was causing the audio to play. Of those, two pairs then spent several minutes interacting with the piece, deliberately triggering responses to signs of amusement. 

Whilst all the above is presented as preliminary observational data, it undoubtedly provides insight into both the effectiveness and reception of the design.  

Haiku Whisper is presented here as an investigative installation for reconnecting both with a diminishing cultural aesthetic and our own urban audio environments. The tool with which it does so consists of a hidden interactive Haiku generator, developed to perform a unique, coherent Haiku to an unsuspecting audience in a public space in reaction to their movement. Inasmuch as the intention was to create both a system and a reaction, it is a success. When analysed further, there is certainly room for further exploration. It was a stated aim for the work to create a sense of belonging in the audience, even though they would originally not be aware of their role in the installation, and this was only partially successful. The nature of being a ‘hidden’ work dictates that a majority of participants would not engage fully with it. Perhaps the next incarnation should find a middle ground between being completely hidden and explicitly displayed. 

  1. Dietze, S. 2002. Ten Dreams of Technology. Leonardo 35.  Vol.5. Pp. 509-522.
  2. Hiraga, M. 1999. ‘“Blending” and an Internpretation of Haiku: A Cognitive Approach’.  Poetics Today. Vol. 20, No.3, Metaphor and Beyond: New Cognitive Developments. Pp. 461-481. 
  3. Manurung, H. 1999. ‘An Evolutionary Algorithm Approach to Poetry Generation’ in Proceedings of 1st International Workshop on Literature in Cognition and Computing (1999).
  4. Olivera, H.O. PoeTryMe: A versatile platform for poetry generation. CISUC ‘08. University of Coimbra, Portugal. 2008.
  5. Ong, W. ‘Orality and literacy: The Technologising of the Word’. London, UK. Routeledge. 1971.
  6. Schloss, W.A. and Stammen, D. ‘Ambient Media in Public Spaces.’ Sophia Antipolis Microelectronics Conference’08. (SAME ’08). Vancouver, British Columbia, Canada. 
  7. Tosa, N. 1998. Interactive Poem. In ACM SIGGRAPH 98 Conference Abstracts and Applications (SIGGRAPH’98). ACM. New York, 300.
  8. Tosa, N.,Obara, H., and Minoh, M. ‘Hitch Haiku’. 2007. Proceedings of the 2nd International Conference on Digital Interactive Media in Entertainment and Arts (DIMEA’07). 6-7. Perth, Australia. 
  9. Turkle, S. 2011. Alone Together: why we expect more from technology and less from each other’. Basic Books, New York.
  10. Zhu, K., Ranasinghe. N., Edrisinghe, C., Noel, O., Fernando, N., and Cheok, A.D.  ‘Poetry Mix-Up’. ACM Computers in Entertainment. Vol. 9, No.2, Article 8. July 2011.
So here is a little run down of my first attempt at playing with interactive multimedia computer stuff. Heard of Arduino? It's a tiny little micro computer that has limitless possibilities. So, basically, having no idea what any of this stuff does, I 'hacked' (a new word in my lexicon, in this context anyway) a broken old Dalek toy I found on a market stall for 50p, and made it sing and dance to my tune. Pointless? Certainly. Fun? definitely. A huge learning curve. Just for the record, they are paying me to do this...

To Find an obsolete or discarded piece of equipment, and repurpose it: ‘The Disco Dalek’

Sasha Scott

My project was defined simply by the potential that could be found in taking apart a discarded children’s toy looking inside. The dilapidated and discarded model of an ‘evil’ creature from a popular TV series was lobotomized, reconditioned and rehabilitated to become a disco-dancing queen. By hacking the old circuit board, adding an elaborate LED array, and a simple bit of code, it was possible to make it move, speak and display itself like a peacock in all its finery.

The most prescient line in the brief for me was ‘complicated is not always better: make something fun, quirky and elegant’. With this in mind, I set out for my cruft. I had decided I would let whatever interesting items I came across define my project as opposed to hatching an idea and searching for objects with which to fulfill it. An afternoon in Deptford flea market proved highly fruitful; I came away with several old, broken and incomplete toys, and all costing less than 50pence. My only criteria for choosing what I did were that I must be intrigued by what I might find inside. On closer inspection a radio-controlled car was dismissed for being too boring and therefore lacking potential, and a toddler’s word-play game was discarded for being in too far a state of disrepair. I was left with a battered but brilliant ‘Dalek’ model. Despite its lifeless state, it contained plenty of interesting parts, including an intriguing audio system.

The Project
First, it was necessary to decipher exactly what was going on inside. A lifeless circuit board, several loose wires and a lot of heavy duty adhesive had to be traversed, and within not much time I realized it had previously been remote controlled. This was good news, as after a few remedial repairs, the radio receiver on the board could easily be bypassed. By simply linking in my Arduino board and giving the two circuits a common ground I now had control of two motors to control movement, and five pins that activated various catchphrases from the famous ‘Dr Who’ series. So now I had to decide what I wanted to do with it. I decided I simply wanted to reverse its character. The Dalek we know from TV is an evil alien creature determined to rule the world through violent suppression, in the form of all the best ‘baddies’ around. So wouldn’t it be great to turn this evil being into a fun loving party animal? If I had control of its movement and speech, this was well within my reach. However, to really be the party, he would need to ditch the drab black outfit as well, and the best way of doing that would be to brighten him up with some kind of light display.

The original circuit board, and the Arduino taking control.

Next came the need to brighten everything up. As Daleks are adorned with hemispheres of black plastic in rigidly ordered fashion, I decided I would base my lights on this. Therefore, 56 individual LEDs were needed, creating my first large project issue. Arduino boards only have 14 pins with which to disseminate information. I have already assigned nine of these to voice and movement controls. My answer would be found in the 74HC595 shift register, a devise that comes allows the control of eight outputs at a time whilst only using three pins on the Arduino board. It does this essentially by pulsing one pin up and down, thus communicating a data byte to the register one bit at a time. A second pin then delineates between those bits, in a process known as "synchronous serial communication". However, 8 pins still do not accommodate 56 LEDs! My answer lay with the ‘Serial OutputPin’ on the shift register, which allows for the information received to flow out again unchanged. This facilitates the ability to transmit 16 bits (2 bytes); the first 8 will flow through the first register into the second register and manifest there. The registers can then be daisy-chained until the number of outputs required is achieved; I needed 7, and the input is therefore multiplied and formatted accordingly.  

The need for 7 shift registers did not in itself present a problem, but along with the requisite circuit board, wiring and LEDs, it became clear very quickly that space in the internal cavity was fast becoming a premium, and as such I would have to be very careful to keep everything as small as possible, employing my fledgling soldering skills to their full capacity.

The shift register circuit before and after securing it inside the model.

The master plan was to set the functions and possibilities through Arduino, and control or trigger them through MaxMSP. The idea was quite simple; by using a beat tracker, when music was played a set of responses would be triggered in the Dalek in time to the music. This idea evolved into mapping three tonal peaks (low, middle and high), and having the three functions respond to separate parts of the music. Unfortunately, time proved too elusive to allow me to fully implement this.
For the movement, I wanted to give some grace to an otherwise clumsy creature. I decided a Waltz was appropriate, and the only classical dance sequence I might aspire to achieve with such limited functions. I was semi successful in this; I mapped a large part of the required steps, but implementing them in a code that flowed in relation to the other functions proved too much. As such, I had to seriously dumb down the movements, but this is certainly an area that has potential for a lot more fun work.

The voice function threw up other issues. My concept was to make the Dalek ‘sing’ of sorts. The functions run by a single switch, trigger by a simple on/off command from the Arduino. Each voice command triggered a full sentence, but if another was triggered quickly it would over-ride the first, allowing for a ‘beat box’ effect. This was achieved, but I was left unsure how apparent it was to the outside audience. Having spent so long with it, to me it was obvious what I had manipulated, but I suspect it was not so for others.
As mentioned above, time proved too elusive to fully implement all the coding I was going for. However, I put together a rather clumsy string of code that at least displayed the various functions that I had either created or taken control of. This threw up its own complications, as all the separate functions had to be interspersed with each other to make sure it flowed as a comprehensive whole. Despite this, I still feel the project was a success. I managed to alter and manipulate the Dalek, and even if I did not use that to its full effect the potential is there to make it do all manner of things; the control mechanisms are all in place.