<p>According to the loose speech theory of knowledge, strictly speaking, knowledge is a very demanding state, one that we rarely occupy; however, knowledge ascriptions are instances of loose speech and thus often convey truths. Just as I can convey a truth by saying, “This class begins at 9 AM,” even if the class starts a trillionth of a second past 9 AM, I can convey a truth by saying, “I know that I have hands,” even if I merely have strongly justified, high confidence that I have hands. This paper is a defense of the loose speech theory. On the negative side, I address a series of objections—first, that we ascribe knowledge in the absence of certainty and infallibility; second, that the loose speech theory is uncharitable; and third, that the loose speech theory delivers an overly demanding interpretation of knowledge norms. On the positive side, I show that the loose speech theory neatly dissolves a difficult problem about inductive knowledge, and I conjecture that it can dissolve many other problems in a similar way.</p>