I first heard about the term “Granular Addressing” from Doug Engelbart. Granular addressing is the ability to address a small fragment of a document. Here is a hypothetical situation.
I am reading an article on venture funding of search startups. I would like to share a few sentences that mention the funding activity to one of my friends. I have a few choices.

1. Copy and paste the fragment of text along with a link to the original article and email it (this is what I do most of the time).
2. Get one of those page annotation services, highlight the portion to draw his attention, add a comment on why I think this fact is important and send an annotation link.
3. Mark the fragment, book mark it using del.icio.us (del.icio.us automatically copies the marked portion to the description of the bookmark) and add it as a shared bookmark.

If Doug’s concepts are implemented, granular addressing may become a feature of the web. See the activity related to HyperScope on the current effort.

The cutting and pasting stuff works good in most cases when the document is text. What do you do when the document is an audio or video file?

Google seems to have a solution for the video case. According to this blog entry, you can mark a portion of a video, optionally specify a fragment duration and share the link.