The Coherence Theory of Truth is probably second or third in popularity to the Correspondence Theory. Originally developed by Hegel and Spinoza, it often seems to be an accurate description of how our conception of truth actually works. Put simply: a belief is true when we are able to incorporate it in an orderly and logical manner into a larger and complex system of beliefs.
Sometimes this seems like an odd way to actually describe truth - after all, a belief can be an inaccurate description of reality and fit in with a larger, complex system of further inaccurate descriptions of reality. According to the Coherence Theory of Truth, that inaccurate belief would still be called "truth." Does that really make any sense?
Truth and Reality
It would help to understand the philosophies of those who defend this theory - remember, a person's conception of truth is deeply intertwined with their conception of reality. For many of the philosophers who argue in defense of the Coherence Theory, they have understood "Ultimate Truth" as the whole of reality. To Spinoza, ultimate truth is the ultimate reality of a rationally ordered system that is God. To Hegel, truth is a rationally integrated system in which everything is contained.
Thus, for system-building philosophers like Spinoza and Hegel, truth isn't actually divorced from reality, but they perceive reality as that which is described in a total, rationalized system. Thus, for a statement to be true it must be one which can be integrated into that system - not just any system, but the system that provides a comprehensive description of all of reality. Sometimes, it is argued that no statement can be known as true unless we also know if it coheres with every other statement in the system - and if that system is supposed to consist of all true statements, then the conclusion is that nothing at all can be known to be true or false.
Truth and Verification
Others have defended a version of the Coherence Theory which argues that true statements are those which can be adequately verified. Now, this may initially sound like it should actually be a version of the Correspondence Theory - after all, what do you verify a statement against if not reality in order to see if it corresponds with reality?
The reason is that not everyone accepts that statements can be verified in isolation. Whenever you test an idea, you are also actually testing a whole set of ideas at the same time. For example, when you pick up a ball in your hand and drop it, it isn't simply our belief about gravity which is tested but also our beliefs about a host of other things, not least of which would be the accuracy of our visual perception.
So, if statements are only tested as part of larger groups, then one might conclude that a statement can be classified as "true" not so much because it can be verified against reality but rather because it could be integrated into a group of complex ideas and they can then be verified against reality. This version of the Coherence Theory can be found most often in scientific circles where ideas about verification and integrating new ideas into established systems occur regularly.
Coherence and Correspondence
Whatever the form taken, it should be clear that the Coherence Theory of Truth isn't that far from the Correspondence Theory of Truth. The reason is that while individual statements may be judged as true or false based upon their ability to cohere with a larger system, it is assumed that that system is one which accurately corresponds to reality.
Because of this, the Coherence Theory does manage to capture something important about the way we actually conceive of truth in our daily lives. It isn't that unusual to dismiss something as false precisely because it fails to cohere with a system of ideas which we are confident are true. Granted, maybe the system we assume to be true is quite a ways off the mark, but so long as it continues to be successful and is capable of slight adjustments in the light of new data, our confidence is reasonable.