What is it? Instrument or ensemble?

Lars Bo Andersen
What is it? Instrument or ensemble?
Humans & IT research seminar

This paper was presented at an internal research seminar at the Humans & IT research programme at Aarhus Univeristy. The purpose was to stimulate debate over IT as an instrument or material for research.

This seminar is our occasion to debate and reflect on IT as an instrument and experiment.

However, this presents me with a problem. Because what do we really mean when we say IT?

It seems to me that this acronym is a blackbox sitting on top of all sorts of very different things, ideas and practices. One has only to appreciate the rich diversity of research projects unfolding in our programme under the common guise of IT.

Perhaps more than a common instrument (or material), IT is an ensemble of theoretical concepts, methodologies, collaborations and practices.

Accordingly, we have become rather good at asking what is IT here, what is it there, how may it fit into this situation, how may it change that situation, and so forth. We know very well that IT is all sorts of different things although it is also only one thing.

So I would argue that IT is not really an instrument, that it is in fact an ensemble or many different ensembles.

However, and this is my point, perhaps we are less inclined to consider our instruments in the same way as we do all of those different IT-ensembles.

Here I am thinking about all those concepts and methodologies through which we constantly dig out new forms of IT.

After 25 years of information- and design studies, perhaps it is time for us to revisit our instruments as if we had to learn them anew.

Turning them into ensembles may help us do just that because then they are no longer instruments.

But what is an ensemble?

First of all, an ensemble is a multiplicity void of any core essence.

Secondly, whereas instruments retain some coherence across time and space, an ensemble is more like an event, it fluctuates and become different in resonance with all those many different things of which it is made.

So we become uncertain. How will the ensemble perform tonight? What will it be this time?

This is rather contrary to the usual function of instruments as proven tools for transforming, manipulating or investigating other things without being affected too much themselves.

After all, I as an ANT theorists can identity thousands of actor-networks knowing pretty well how my conceptual instruments will perform in each case.

The point here is difficult. Importantly, it is not that we are not developing our theories or methods.

On the contrary, the point is that turning them into ensembles render them receptive for all sorts of new connections and disturbances – some of which are already implicitly there awaiting only to be recognised.

Deleuze would say we are looking for lines of flight, Latour would say we are looking to establish novel translations, but we could also just say that we are revisiting the instruments with which we work in order to make them uncertain again.

As Foucualt once said about his own instruments: They're meant to be dissolved!

What is participation?

So I will now ask what is participation and forbid myself to consider it an instrument or method.

Participation has been central to much of our research for the past 25 years.

However, while it is clearly acknowledged that participation in practice is highly diverse and complex there is apparently only limited theoretical interest in asking what is participation - it is too much an instrument and too little an ensemble.

Luckily, participation has recently been taken up anew by various people amongst us.

For instance, Peter Nielsen is doing an entire PhD asking what is participation and Winnie Soon is asking the same to programming languages and APIs.

But what I will say now is based on a collaboration with Peter Lauritsen, Peter Danholt, Kim Halskov and Nicolai Brodersen.

Together, we have written an article with a simple agenda: to make participation an ensemble.

We have chosen ANT to do so.

Put very shortly, ANT is basically the claim that the world consists of ‘actor-networks’ continuously transforming and composing other actor-networks. If they touch, they change, if they exist, they exist as networks of many different things.

Consequently, participation is rooted not in subjects but mediated in actor-networks composing and transforming each other.

So ANT steals away the user and gives us a network in its place.

The consequence being that participation eludes participatory activities, it is no longer contained by them, and comes into play all over and often in rather surprising situations.

I have to short illustrations:

The first:

Since 2013 we have been working on a design project called Teledialogue.

The purpose of Teledialogue is to use participatory methods to design some form of IT that will enable children placed in foster care or at institutions to have more frequent contact with their social worker at the municipality.

Teledialogue was, amongst other, motivated by a report by the National Council for Children (NCC) in Denmark.

The report is based on interviews with 113 placed children and provides a collection of suggestions on how to improve their situation.

One thing these children wanted was improved dialogue with social workers and to have social workers include them more in decisions affecting their own life.

While never physically present in our project, these 113 children gave Teledialogue its purpose of improving communication with social workers through IT.

Their frustrations with the current social system were inscribed in our funding proposals which eventually went through.

So presently, when we invite other children for workshops, our actual users, whatever they come up with is overtaken and premised by a report with 113 other children.

It does not matter if these children would rather improve relations with their pedagogues or design something for their friends because the report children are always participating with them, directing our design towards social workers.

The second:

In Teledialogue, children share personal and intimate information with social workers. Participants may thus be concerned about privacy and data security.

However, this concern evades children directly and is dispersed instead onto system administrators, legal advisors and most importantly the Danish Data Authority (DDA).

Since children are talking to social workers (that is, representatives of the state) management of privacy is inescapably interfered with by the DDA.

To stay with the musical metaphors, it does not matter which tune the children play, the DDA will always be part of the act when they perform.

For instance, many placed children use Skype to keep in contact with their biological family. It goes without saying that they are talking about sensitive and private matters. However, the children are happy with this and would like to use Skype to talk to social workers too.

But in their capacity of being placed children talking to social workers, their stance towards Skype is overtaken by the DDA who are not convinced that Skype is sufficiently secure (the same applies for Facebook and other popular social media).

Whenever there are placed children there is always also the DDA.

Participation as an ensemble

Who are speaking on behalf of who and who are speaking only for themselves? Whereas instruments would give us users, an ensemble provides neither individuals nor trustworthy representatives.

For instance, one could speculate if I am speaking here as individual or representing others?

It is not possible to know what a placed child unrelated to anything and undisturbed by others would think of Teledialogue.

ANT, in fact, raise doubts to the very existence of children who are not ensembles.

In our article we have tried to emphasise this through two analytical notions:

  1. Participation as overtaken; meaning that agency emerges from many sources and is not easily locatable.

  2. Participation as partially existing; meaning that different forms of participation unfolds throughout a design project, even when writing funding proposals.

As such, participation does not come into full existence at workshops only to disappear completely when writing articles or drafting research proposals.

If participation is considered an ensemble rather than an instrument, it becomes an event and achievement in each situation and thus something we do not already know or control.

Instead, we are given the opportunity to stubbornly trace through the ways in which children, funding proposals, reports and paperwork, legalities, coordinating meetings, boyfriends and parents are interrelated trying somehow to influence and overtake all those many different elements of the ensemble.

Of course, such efforts are often futile, we only ever know half the ensemble and even those we do know may easily evade us.

In the words of Foucault, participation becomes a set of actions upon other actions - drifting along with us trying to hook in at various stages.

In conclusion

The argument can be summarised as follows:

Instruments are somewhat coherent entities, it is only the IT-world around them that is many different things at once.

Here we constantly ask 'what is it' anew even when we have been researching the same phenomena for 25 years.

We treat IT as an ensemble and not as an instrument.

So what would happen if we treated our methods and concepts the same way? As ensembles which we never really know until they unfold in performance?

In relation to participation my fellow authors and I found three things:

1) Participants are not stand-alone subjects but network configurations.

2) Participation is not limited to designated events but always partially at play.

3) There is no gold standard of participation, only an imperative to account and investigate.

So the question is, how can we further allow our instruments to become dissolved in each specific ensemble, in each performance and where does that leave us?