Babelona; The tower of
6 October 5th, 2010 at 06:10
Papua New Guinea is home to half of the world’s languages, which probably explains why its capital is the mysterious sounding “Port Moresby”. Barcelona last week became the capital of learning object metadata languages, as an international gathering of repository people came together to discuss new ways of sharing learning objects and their associated metadata. Whereas Papua New Guinea is home to over 4000 languages, it turns out in metadataland we are well on our way with over 130 already. So what looked like it would be a conference about ways to share learning objects, and maximising reuse, soon became a series of PowerPoint translations of the opening credits of Dad’s Army. DCMI did attack LOM, and LOM did attack OAI, and before you knew it a Lilliputian metadata civil war started. Please, warmongers, won’t someone think of the end user experience?
So, thought of harmonisation and finding a common tongue across all this data came to the fore. However, hopes of performing a “reverse Babel” and backtracking into one great repository language did sadly end up as a “despairanto” – a complete non-language that didn’t really seem to offer anything of any great use. The oft-mentioned “silos of data” appears to given birth to “silos of languages” where the data is so tied to formats of representation that the data isn’t really data anymore. The desire to present metadata in one format, has led to the data itself becoming almost a monkey to the representational requirements of the organ grinder, and boy, is there a lot to grind.
Controlled vocabularies came up a lot. Repository users could have the chance to add metadata to the repository, but only if it conformed to a list of words chosen by the repository manager. I can only imagine the poor people of the Papua New Guinean provinces of Simbu, Western Highlands and Madang when they come to catalogue Mount Wilhelm (or in German, Wilhelmsburg, or Enduwa Kombuglu in the local Kuman language). How can the controlled vocabulary deal with that? If you take one thing from this blog, please, do think of Papua New Guinean repository managers and the problems they must face.
Speaking a little repository language, I could just understand most of the repository-speak (and when I didn’t I just asked them to shout it louder and slower), but not the thinking as to why these difficulties are seen as so pressing to need to be discussed. No one seemed to wonder about the end user experience, or theories or thinking on how to best expose these learning objects to people so people could actually learn from them. Progressing from my biblical Babel I started to wonder about repositories, and which repository came first – then, eureka – Noah’s ark. Noah obviously keeps LockSS fans happy by having two of everything, and he has one of every living thing on the planet (if you ignore unicorns, but let’s not cry over horses with experimental milliners) – so the only criticism you could level at Noah was that he probably didn’t conform to any data standard, but then he did store two of every living thing in the world.
So perhaps from Noah we can take his Ark ideology (sadly Noah didn’t open source his boat plans), and apply a simple and light framework to storing learning objects.
A) – Acquire content in as many ways of possible, taking as much metadata as can be provided by a user, and then automatically generating as much as possible.
R) – Reuse, simplify and facilitate as much reuse as possible. Allow for crowdsourced metadata and metadata generated via reuse. Allow reuse to create didactic metadata flows.
K) Know about it – allow the knowledge generated through acquisation and reuse to inform and guide metadata, and community development around learning objects.
I’ve started on the code for this already, I will send a dove for you when I am done.