Although the concepts of truth and belief are fundamental in philosophy, in recent years they have come under attack from various quarters. I argue that philosophers have been too quick to find these concepts problematic and in need of being replaced. For example, the Liar Paradox is sometimes taken to show that the concept of truth is inconsistent, and thus unsuitable for rigorous inquiry. But I develop a solution that gives a consistent account of this concept, allowing us to retain it in spite of the paradox. I argue that when the word ‘true’ occurs in such a sentence, it undergoes a one-off aberration in its reference, failing to refer to truth. Thus, Liar sentences and their kin fail to say what they pre-theoretically appear to say. However, there is no need to conclude that these sentences are meaningless; rather, I illustrate how these sentences come very close to saying what they appear to say, in spite of the aberrations they witness. As with the concept of truth, intentional concepts like those of belief, meaning, and reference have been subject to skepticism and attempts at excision. I show that the content of the claim that the use of intentional concepts can be eliminated from scientific explanations depends on broader issues about how one conceives of explanations. Then I argue that intentional concepts, in particular the concept of belief, play an ineliminable role in the explanation of behavior: when we learn what someone believes, we get some information about how she would react to a variety of possible scenarios. This information that is useful in everyday life, and would be important in a science whose aim was to improve on our folk-psychological explanations. But so far, explanations that avoid talk of beliefs have failed to replicate this distinctive kind of informativeness. Thus, we have reason to think that belief-attributions are indispensable when it comes to explaining people’s behaviors, and so we have reason to retain the notion of belief.