Theory of Poker has a little about it in it's "Reasons why you raise" section. SSH has a bit about the cheap showdown that has concepts that apply as well.

Basically, if you're committed to calling a river bet because your hand is good enough it may well win, then raising on the turn may get you there for the same price, but has a couple extra bonuses:

1) The bettor may fold to your turn raise figuring he's badly beat.
2) You may draw into a great hand on the river that you can then bet and get and extra BB out of the hand.
3) If the board has a draw, like a flush draw, you punish them for an extra BB when drawing. A BB you won't get on the turn because they won't bet if they missed their draw.

Circumstances have to be fairly specific though.
1) You should have a decent likelihood of improving on the river, such as 5 outs to 2-pair/trips. If you don't have a real shot of improving, just call the turn and river and get to showdown. A 2-outer is not a "decent likelihood" [img]/images/graemlins/smile.gif[/img]
1a corrallary) Improving your hand won't make the board god-awful scary (like a 4-flush).
2) The opponent needs to be someone you can cow into passively calling your turn raise and checking to you on the river.
3) You have to have reasonable odds of having the best hand now. Say you have KdTd on a board that's AdTs8s3h, that's a good candidate (unless the read on the villian indicates he MUST have the ace, like a 1% PFR or something). The villian could be on a draw, could have QTs, maybe 99. Reads apply of course.
4) The villian is likely to call your river bet when you improve. (ie. what's his WTSD%?)

I like the play, but I don't use it frequently simply because the circumstances don't come up frequently. To some extent it's like the free card play. Raising for a free card isn't a good idea if you're not going to actually get one. Same thing with the free showdown, raising for it isn't a good idea unless you meet the conditions above.