Categories of semantics

Whenever you choose an HTML element to mark up your content, your choice will fall into one of these categories:

1. Element chosen is valid, semantically correct, and structural.
2. Element chosen is indisputably wrong according to the validator (e.g., div inside p)
3. Element chosen is indisputably wrong according to a plain, uncontested reading of the spec, even if the validator passes it (e.g., using div instead of p)
4. Element chosen is presentational when a structural element could be used (e.g., tt for samp)
5. Element chosen is structural but could be replaced with a structural element with better semantics (e.g., p should really be address)
6. Presentational element is chosen because the content does not suit itself to a structural alternative (e.g., b and i in marking up historical documents)
7. Element chosen is semantic but relies on sometimes-disputed interpretation of the spec (e.g., definition lists used to mark up dialogue)
8. Element chosen is semantic but pushes the very limits of its definition in the spec (e.g., individual Weblog entries marked up as li)
9. Element chosen is generic (i.e., span or div) because no other element conceivably fits

Please make your selection now.