I think that this assumes that something like 'eternal knowledge' (and implicitly, 'truth') exists, which I think is the fundamental problem with this reasoning. The best we can probably hope for is to agree on some definition of what consistutes truth and then strive to achieve that.
Science is the best example of this- the validity of predictions (or the failure to falsify predictions, more precisely) is what constitutes truth in the community of scientists, and as our ability to experiment and measure grows, whole sections of science are discarded and replaced with better explanations.
From: Joho the Blog
A couple of days ago, I wrote up a thought that I was afraid sounds better than it is. But now I think maybe it isn't as hollow as I'd thought. The idea was this: Aristotle says that to know x is to place x into a relationship of similarity and …
Link to article...