Whose voice?

Another too long day with an early morning flight to Amsterdam.   I had booked an airport hotel and should have been there by midnight.  However a fatality on the line at Battersea Park delayed all trains so I had another night with sleep restricted to four hours.  By the time I made it back to Lockeridge that evening it took an effort (worthwhile) to cross the road to the Who’da Thought It for a pre-christmas drink with friends.  I was there for a Development Sector seminar in support of our new Benelux operation formed with Dave and Friso.  The first of many such partnerships it should also us to scale across geographies and applications as the market for sense-making grows.

One of the key principles of SenseMaker® is to allow the voice of the subject to be heard, not just as a narrative but also as interpretation.  I normally contrast this with longitudinal research and computer based search (such as sentiment analysis).  I do so not to denigrate the former but to differentiate what we do from them.  In one approach to longitudinal research narrative is gathered in the field, transcribed then tagged or interpreted by the researcher using some theory of the field.  With computer search engines the text is interpreted and patterns sought.

With SenseMaker® we take a different approach.  The theory that the researcher uses is taken and represented as triads and more recently stones but lets keep this simple for the moment.  To understand this process its worth reading Beth’s paper which explains how we went through the process with cultural anthropology to create a signifier set which has stood the test of time in multiple projects over three continents.  Key to this patented approach is:

1. The need to trigger the novelty receptive capability of the brain by exciting its curiosity and by preventing the gifting and gaming that is an inevitable consequence of the classic survey questions where you know what answer the researcher desires.
2. To facilitate this the labels need to be all positive or all negative
3. Given that the primary meaning is in the signification, we can allow people to withhold their story, something that we do with permission codes.  This is important as an issue of respect, let alone its utility for research itself.
4. It also allows us to capture material in picture or sound form which is often more expressive than text and allows people to use their own voice in their own language or dialect.  Given that the primary meaning is quantitative, the explanatory narrative will only need to be translated or seen to provide an explanation for a statistical pattern.
5. Critically it also means we can move to continuous capture in what was once called by a Government Anthropologist the first example of distributed ethnography.

There are other aspects as well, including some our recent work on human sensor networks, but that is for another day.  The point I want to emphasise in this is that the primary interpretation is done by the person who created the story not by an expert.  Secondary interpretation can be done by said experts, or by the people who create them but we start with numerical data created in the field.

I should also say that this process seems to add meaning rather than assuming the story is the be-all and end all of meaning.  A problem for computer based interpretation over and above the wider issue of the degree to which language changes meaning in context, in all too subtle ways for an algorithm to understand other than in the round.

The ideological point however is paramount, ownership can all too often be abducted through the process of interpretation, we seek to return that to the subjects.  Along with that is the idea of peer to peer knowledge flow, something I will pick up tomorrow.