I've been reading Matt Cutt's guidelines on URL canonicalization:

(URL address blocked: See forum rules)

I understand the principle behind this but I don't understand whether the actual content of the pages plays any part in Google being able to resolve them as the same.

For example, I have two pages with two very similar URLs that Google has resolved as being the same page. Both of these pages have very similar content. If I change the content of one of these two pages so it is substantially different than the other page, will this have any bearing on whether Google is able to resolve them as the same or is it simply URL based?

Any thoughts, opinions or guidance would be really appreciated. Thanks!