One of the things that established researchers have is a working knowledge of the relevant literature. Most established researchers have a core repertoire of some 100-150 works on which they can draw readily. These are a useful selection from the hundreds or thousands of articles and books the researcher has digested over time. The repertoire gives a researcher a context in which to place ideas: the collection characterises the major strands of thinking in the field, identifies the major researchers, and provides research models and examples. Of course the repertoire evolves and must be updated.
Part of doctoral study is acquiring one's own core repertoire. The annotated bibliography is an effective mechanism for facilitating this acquisition and for keeping record of the majority of papers that fall outside the core. The annotated bibliography is a powerful research tool. It should be a personal tool, keying into the way you think about and classify things.
It should include, as a minimum:
It can include many other useful things, e.g.:
Keeping an annotated bibliography is a discipline. It is easiest to establish a discipline of writing notes about papers as soon as you read them, not going on to the next paper until you have done so. It's much harder to go back and try to catch up. Because keeping the bibliography is an `overhead', and because the point is to maintain access to material, it's best to keep entries to under a page per paper.
Never delete things from the bibliography. `Discards' can be re-categorised or filed away separately, but one year's `junk' may be another year's `gem' (and vice versa). There is also genuine value in keeping track of the changes in categorisation: one way is to keep a list of working category `definitions'. Don't discard the old scheme after a re-vamp, rather file it as part of the record.
The discipline is to keep up a continual, accumulating record of your reading and thinking.
The bibliography can help you to `backtrack' on your own thinking.
It will reflect the evolution of your reading, of what you found important over time, and of your writing about what you read.
When you find a reference and can't remember the paper's particular perspective, the notes can give you the key.
When you re-read a paper just before your viva and say: "Oh no, it doesn't say that at all, what could I have been thinking?", the notes will be invaluable.
The bibliography can help you to manage your reading effectively and keep accessible much more information than you can remember without aid.
Keeping a bibliography allows you to use a `flat', unambiguous physical filing system (e.g., alphabetical by author) while being able to categorise, re-categorise, and search fluidly.
The bibliography can help you avoid re-reading papers that are useless and forgettable but have interesting titles.
It can help you keep track of the physical form and location of materials.
There are different ways to keep a bibliography. The most common forms are card catalogues and electronic databases.
Papyrus http://www.rsd.com/ ProCite http://www.risinc.com/ EndNote http://www.niles.com/
Many people don't use specialist packages, preferring to adapt database or spreadsheet or word processor usage. Many effective bibliographies are simply kept as very long text files.