Taggs, please do continue to present your point of view; I find that the most valuable part of almost all really useful articles here are the discussions!

In this case, as Gethyn says, there's more than one way, each with benefits and disadvantages.

1) You can stop taking backups after the first sign of corruption; if you immediately notice that, and immediately correct it, and your prior backups are still available, this may well be the best choice.

2) You can continue taking backups even after the first sign of corruption; if you do not immediately notice that, particularly if you keep backups for a long time, you later have more choices on backups to restore. If you don't keep backups for a long time, especially if days or weeks of business transactions have occurred since the first sign of corruption, and the corruption is easy to handle (oh... we don't care _that_ much about that index/table/data anyway!), then you can at least still execute a normal restore for other reasons (that disk that caused corruption completely failed (as did another in the raidgroup), etc.), and deal with the corruption later, rather than tell the business "Well, we can restore... to five weeks ago."