« Moving CoPs ahead | Main | Questions matter »

November 11, 2003

Comments

David Locke

I've written glossaries professionally. I have not moved to thesarus construction, or controlled vocabularies. Glossaries fit into the controlled vocabulary continuum as well as taxonomies and classification.

Definitions serve as the locus for domain-based culture. Being on or in a domain are different orientations based on focus. Am I surface, or am I a critical expert, or am I a driving expert. The definitions of these individuals are not the same. What does it take to make the cultural shift?

Definitions are also conceptualizations. They provide a place for a concept to exist and be communicated. Finding the definition is a communications task that happens after the concept is discovered, after Aha! The definition is the means to market the concept. And, in a market a concept has many different definitions.

Definitions are taxonomies. They are the taxons that allow use to say what differentiates this concept from others. Definitions are likewise ontological, because the taxons can only arise, if they are sortables within the ontology. Can I see it, can I count it? If not I can hardly define it. Ontologies determine who is in and who is on in regards to a specific discipline.

So a defintion arises in taxonomic, ontological, and social networks. Likewise, they participate in processes that tie these networks together: concept creation, conceptualization, bibliographic maturity, adoption by the research community, adoption by markets.

Defintions exist in temporal windows, because the taxonomies and taxons, the selection criteria are always under assalt by competitors. Each seller has a definition that is slightly different from the other sellers. Lexical warfare in the market and corpi erode defintions to the point of meaninglessness.

Concepts arise either divergently or convergently. In the divergent case, you are starting with on corpus. In the convergent case, you are starting with multipile corpi. There is no one answer for bootstrapping. My on take on bootstraping is that everyone involved brings their own personal definitions and taxonomies to the table.

With taxonomies, this means post-coordination. It means accepting the fact that there is not one taxonomy or one definition at the beginning. The effort is to coverge and bridge the taxonomies in a way that doesn't reduce diversity.

This also means that there is no power authority. The library science definition of authority means preferred term and not who has the ability to add, suggest, change, or link. I've worked in many places where I didn't have a power relationship, so I couldn't participate in and defining. I think that power authority is destructive to social construction.

The market effects on terminology, definition, glossary, taxonomy, and ontology work against any notion of a central, power-authorized meaning. Your bosses competitors can always undermine his conceptualization.

All these definitions and taxonomies create a multitude of links. Information architecture organizes content (nodes) and links. The preference today is away from inline links. This plays with the template notion. I use information architecture frameworks in my work. I resist inline links, instead I use a consistent layout that says if you need more information on a facet look here. And, that is where I list the links. Each facet gets its own list of links. Consistency sets expectations. Consistency meets expectations.

Keep in mind also that in hypertext theory nodes and links have equal weight in defining knowledge. A link is not a page turning device. A link is an association between the source and the target content. A link can be computationally complex as well. A single link might open all the competing definitions rather than just one.

Level of abstraction is a continuum. You should be able to move from corpus entry to summary to conceptualization to concept to keyword (instance) to keyword (class or classes). The corpus entry is the definiton. It is the text that should appear in the glossary. A glossary should not use any terms not in the corpus entry.

This means that, while glossary construction is valuable, it is the corpus entry is the act of definition. Glossary is a post-creation, metadata process. Collaboration should stay focused on creating the written definition within the corpus document(s), then glossary construction will be straightforward and will not involve power authority.


The comments to this entry are closed.