A recent post by Andrew Townley, Mechanisms for expressing ‘aboutness’ effectively? made me reconsider dc:subject.
Or perhaps more accurately, what does dc:subject, and similar properties, lack?
It has a certain flatness that can best be illustrated by evaluating it using the 5 W’s (and one H). As listed by Wikipedia (Five Ws):
- Who is it about?
- What happened (what’s the story)?
- Where did it take place?
- When did it take place?
- Why did it happen?
- How did it happen?
The who is answered, but only by attachment to a particular item.
And it can be argued that dc:subject answers what by its value. A simple string value but that is a question of the degree of usefulness of an answer.
But what of where, when, why and how?
They go unanswered.
True enough, that is information about the subject being assigned.
To use misleading terminology poorly, metadata about metadata.
But knowing why a particular dc:subject property value was assigned to an item could help with consistent use of that property value.
Or even discover when other dc:subject values were being used for the same subject.
Just as knowing when a subject was assigned to an item could be used to establish chronologies of subject classification as applied to particular items.
Or knowing where, both as in geographic as well as institutional location may reveal differences in subject assignment for the same items.
Data (by which I encompass the misnomer “metadata”) that lacks the means to answer the 5 Ws (and one H) for itself is impoverished, and unnecessarily so.