Five Ways of Looking at a Thesis[Page 1]
1. A thesis says something a little strange.
Consider the following examples:
A: By telling the story of Westley and Buttercup's triumph over evil, The Princess Bride affirms the power of true love.
B: Although the main plot of The Princess Bride rests on the natural power of true love, an examination of the way that fighting sticks--baseball bats, tree branches, and swords--link the frame story to the romance plot suggests that the grandson is being trained in true love, that love is not natural but socialized.
I would argue that both of these statements are perfectly correct, but they are not both strange. Only the second one says something, well, weird. Weird is good. Sentence A encourages the paper to produce precisely the evidence that The Princess Bride presents explicitly; sentence B ensures that the paper will talk about something new.
Romeo and Juliet concerns the dangers of family pride, Frankenstein the dangers of taking science too far. Yup. How can you make those things unusual?
Good papers go out on a limb. They avoid ugly falls by reinforcing the limb with carefully chosen evidence and rigorous argumentation.