validity

Alan Dix, Lancaster University, UK

Position paper for CHI2004 Workshop on Reflexive HCI
download paper (PDF, 29K)

 

Full reference
Alan Dix (2004). validity. In Reflexive HCI: Towards a Critical Technical Practice, Workshop at CHI 2004, Vienna, Austria 26th April 2004.
http://www.hcibook.com/alan/ papers/chi2004-lit/
More:
Download paper (PDF, 29K)
Previous papers on literary arts and HCI: called xxxxx whilst waiting for a title (2003) and exploring the future in story and folktale (2002)
Alan's essays on imagination and rationality and words page
.:thePooch:. ... lots of fun work in performance arts
web site

innocence of snowdrops beneath rain damp trees
and gold glow summer misted hills,
the sound of laughter in the street outside,
ring hollow in those dry eyes

hands that fed turn magazines and open tins of soup,
feel velvet moss on autumn walks and gather empty milk bottles at night,
but still recall the paper touch and sickly drag of weightless flesh

the sound of jumbo jet or wagner at full volume cannot drown the silent stare,
accepting little when all is lost
and ears still hear the sound of those dry eyes

alan words 2004

dry eyes

Twenty years on from the Ethiopian famine of 1984 the BBC broadcast a documentary. Michael Buerk revisited the people he had interviewed and the places he had visited. The programme also covered the subsequent events, including the Band Aid Christmas record and Live Aid concert.<1>

The scenes from the original footage were harrowing, even knowing that the events took place ago 20 years ago. In 1984 a young woman, Claire Bertschinger, had been in charge of a feeding centre for children at one of the major camps. Each day she would walk down the line of listless shrunken children and babies and select those who would be let inside. She did not choose the weakest. they would be fed for a day and die regardless ­ a day's food wasted. She chose only those she believed would be saved. A few moments assessment of pitted eyes, parchment skin and stick-like limbs and a choice: life or death.

She too was brought back to Ethiopia. For twenty years she had held the horror of that time, and believed that those she had worked with and those that had been at the camp, would regard her, like the commandant of Auschwitz, a dealer in death. Of course she was greeted with joy and love by all who had known her all those years back; they remembered the life she gave. Twenty years on her healing could begin.

Holding our own work against the lamp of these events is perhaps too revealing, but perhaps Band Aid itself is easier to deal with.

So, let's ask the question "was Band Aid a success?" and see what that tells us about the measures of validity in our own arenas.

 

success

Of course in true deconstructive style, Band Aid itself is not an isolated incident, but is part of this wider unfolding. Bob Geldof was moved by the reporting of Michael Buerk, by the images of death and suffering. He was a pop star, and a fading pop star, but also one who, through character or experience, was able to recognise that.

In the documentary he recalls how he thought that a Christmas record could net seventy two thousand pounds<2> but knew that a Christmas record by him would not be the success it needed to be. In humility uncharacteristic of the industry he went instead to his friends and contacts rather than going alone.

Of course we all know how Band Aid was in fact a great success and netted over eight million pounds with its own chartered boats and lorry conveys taking medical supplies and food to the heart of a war torn and drought stricken country.

Now this is a metric of success or validity that a traditional HCI practitioner would love: £8 million raised with an initial specification target of £72 thousand. We might question the professional competency of a designer whose system outperforms expectation by 10,000%, but we can hardly question the success. Hard numbers ­ yes!

But Band Aid was not just a fund raiser, but also a record and a song, so what about aesthetics? Bob Geldof said it didn't matter whether it was a good song or not, he just wanted it to sell. But would public guilt and a hall of fame as singers been enough on its own to make it a commercial success? Could bad music or bad lyrics have been simply a way of drowning out the silent eyes staring from those news reports? I would guess that unless the tune and the performance had been adequate it would not be the case. It is no good it simply being a good record to buy it needed to be a good record to listen to as well.

Of course, now we are treading the ground of more artistic judgement, or at least popular taste.

Not only were the listeners of Band Aid moved aesthetically but also they were often moved to empathy and action. For one Christmas, Band Aid changed the spirit of the public and, because of this, governments also had to change their policy. The £8 million the record made was magnified many fold in state aid. Perhaps the most major effect of the record was its affect.

... and for Bob Geldof himself, his life was changed forever. While the Boomtown Rats are known by one musical generation, his association with Band Aid and Live Aid cut across age and class.

 

for us

The practical success in raising money was an important criteria of success, but of course that required other forms of success. This is exactly the situation facing those of us in HCI as we consider issues of user experience. Like Band Aid neither purely functional nor purely aesthetic considerations are sufficient to understand the full issues.

Traditional HCI takes its notions of theoretical validity from base disciplines such as psychology and ergonomics. Usability testing and metrics have formalised this in terms of measurable efficiency and effectiveness and research looks towards scientific experimental method. The truth of a design rule and the measure of an interface’s value are taken from the external aspects of its behaviour in use.

In contrast, literary and artistic theory looks for its validity in less objective areas. M.H Abrams, in the introduction of The Mirror and the Lamp, says "A good critical theory has its own kind of validity. The criterion is not the scientific verifiability of its single propositions, but the scope, precision, and coherence of the insights that it yields into the properties of single works of art and the adequacy with which it accounts for diverse kinds of art."<3>

It is interesting that this focus on insight does not refer at all to the success or quality of the works being studied. However, it is clear that much of the historic study of arts has focused on trying to understand what it is that makes a work 'good' and how to achieve this quality in practice. For example, the study of metre and rhyme can be used both to account for some of the 'quality' of verse and to guide the poet.

This tension between the singularity of each work and general rules has been recognised for many years. The classical Roman author of On the Sublime, known as Longinus, writes of other contemporary critics "Works of natural genius are spoilt, they believe, are indeed utterly debased, when they are reduced to the bare bones of rules and systems.", and he then goes on to counter this view and produces a work of literary criticism that has been influential for nearly 2000 years!<4>

This classical voice sounds strikingly familiar if we consider those advocates of various models and theories of HCI who seek for generalisable knowledge and those who emphasise the more contextual and singular aspects of each interaction and situation.<5>

This has been a problem with 'normal' work-based systems and the design of them. However, it is even more problematic when the systems we design are intended to elicit emotions, to be fun, to yield experiences. These things take their validity from their subjectivity.

John Searle, famous for his Chinese Room Argument, distinguishes two types of subjectivity: epistemic and ontological.<6> A statement such as "I think the Empire State building is 1273 feet tall" is epistemicly subjective - it is a matter of belief. In analysis, science prefers epistemic objectivity ­ the measured height of the building. However where personal preference, aesthetics, pain or other feelings are the domain of discourse, as in this work, then we have ontological subjectivity - where the subjectivity is the very essence of the thing being studied.

However, the development of certain types of art in the latter half of the 20th century also gives us cause for caution. The subjectivity of experience is transmuted into a critical tradition where the values are aesthetic have no grounding outside the cognoscenti. The role of theoretical critique of human experience should be to explain the felt effects not define what they should be.

 

the real thing

To discuss critical theories M.H. Abrams uses a framework that is surprisingly similar to ones we see in HCI. He looks at four elements: the artistic work itself, the artist who produces the work, the audience for whom the work is produced and the 'universe' the people, events and topics that the work is about. Adams uses this to discuss different critical theories which often tend to focus on one or other element.

In HCI we can substitute designer for artist, user for audience, context and domain for universe and the designed system for artistic work. In HCI we also find that different techniques focus on designer, user or context.

The universe is interesting as there are often two things that a work is 'about' ­ its subject matter and the deeper reasons for its production. Miller's 'The Crucible' is about the witch trials in Salem in 1692 and also ‘about’ McCarthyism in the 1950s.

When Claire Bertschinger first heard Band Aid over a crackling radio set she thought it was a sick joke, someone making money off the back of those she saw dying around her. It was only later that she realised that it was not only about the famine, but 'about' making money to alleviate it.

The first 'about' is concerned with the internal nature of the song. The second 'about' is concerned with the why, the external meaning; it is a single utterance within a wider context: the economics of EU grain and butter mountains, the politics of a war over parched land and dying children.<7>

Band Aid's validity was not in measurable profit or felt experience but in this other 'about'. Looking towards this other validity seems equally important as we practice and theorise in HCI.

 

notes

1. Ethiopia: a Journey with Michael Buerk ­ This World. Clifford Betsall (director), Kern O'Conner (editor). BBC2 9pm, Sunday 11th January 2004.
2. I'm not sure where the figure of £72,000 came from, perhaps a previous Christmas record. However, this is figure the target figure that was iterated several times in the programme.
3. M.H Abrams. Chapter 1 Orientation of critical theories, from The Mirror and the Lamp: romantic theory and the critical tradition. 1953.
4. Longinus On the Sublime. In T.S. Dorsch (trans.) Aristotle Horace Longinus: Classical Literary Criticism. Penguin Books, 1965
5. This tension is also evident in more recent theory. The concept of iterability is central to the distinction and lack of distinction between spoken word and text in Derrida's writings and followers. Each utterance of a symbol creates a new instantiation which is both the same and different from previous utterances: "the structure of iteration ... implies both identity and difference" Jacques Derrida, Limited Inc. Evanston, Illinois: Northwestern University Press, 1988, pp. 53. quoted in Section 4.1 Iterability of Kevin Halion. Speech Act Theory and Deconstruction: A Defence of the Distinction between Normal and Parasitic Speech Acts. PhD Dissertation, McMaster University, 1989
6. J. Searle.The Mystery of Consciousness. Granta, London, 1997.
7. On reading a previous draft of this, monica schraefel pointed out that there was an interesting further context in that this was all happening at the time of the British miners’ strike. Reflecting on this, although there were not millions of children dying in the UK, there was certainly considerable hardship and hunger amongst the families of miners. Furthermore the causes of the Ethiopian famine were as much to do with the war raging there as the drought, and the famine, by depopulating the rebel territories, was to some extent a tool of war. Similarly in the UK the government sequestered the funds of the miners union cutting strike pay and thus forcing many to break the strike due to hunger. monica wondered whether in a way it was easier for people to care about those far away than those on their doorstep. Similarly in preparing this paper I wondered am I, like Claire Bertschinger thought Band Aid was when she heard record, in some way exploiting or trivializing the Ethiopian famine in drawing the analogy between Band Aid and HCI design. Turning this on its head I wonder whether it is more that we can easily as researchers, artists or designers deflect away our own responsibility in the small for our creations by seeing it as unimportant in the large


Alan Dix 28/2/2004