Consider the following examples:

A: By telling a powerful story of failed love, Romeo and Juliet demonstrates the destructive effects of family pride.

B: Although we are told from the beginning that it is a tale of "star-crossed lovers," Romeo and Juliet produces its tragedy by calling attention to a series of near-misses, places where the protagonists' downfall could be avoided.

C. Mercutio might seem like a minor character in Romeo and Juliet, but his language actually tells us something important about how the play works.

I would argue that all of these statements are perfectly correct, but they are not all strange. The first one says something obvious about the play; it is more of a moral than a thesis. Since the play tells us repeatedly about "the destructive effects of family pride," your reader does not need a paper to point that fact out. The other two say something, well, weird. Weird is good. When you start to construct a thesis, think about what an easy one-sentence summary of the text would look like. Then try to come up with something more specific than that, something with a specific twist on the standard interpretation.

This point often separates the best theses from the pack. If your thesis leads to a paper that simply follows the plot of a text as it goes along, it probably needs to provide a stronger statement of the paper's logic. Example A, for example, invites a plot summary of the play. Examples B and C say something more interesting and specific, but you could revise them to indicate more strongly how each paper will build its case.

The MTS: By looking at _____, we can see _____, which most readers don't see; this is important because _____.

Try it out with the above examples. Notice that the MTS adds a new dimension to point number one above. The first part of the MTS asks you to find something strange ("which most readers don't see"), and the second part asks you to think about the importance of the strangeness. Thesis A would not work at all in the MTS. Theses B and C would work only partially; both of them would require some additional information to work better.

If your thesis could describe many works equally well, it needs to be more specific. Let's return to our examples from above:

A: By telling a powerful story of failed love, Romeo and Juliet demonstrates the destructive effects of family pride.

B: Although we are told from the beginning that it is a tale of "star-crossed lovers," Romeo and Juliet produces its tragedy by calling attention to a series of near-misses, places where the protagonists' downfall could be avoided.

C. Mercutio might seem like a minor character in Romeo and Juliet, but his language actually tells us something important about how the play works.

The worst offender by this standard would be a thesis such as "Romeo and Juliet is a powerful story of failed love." There are thousands of stories of failed love in print or on screen, and you can probably think of a number of examples quickly. Example A above is a little more specific, but not enough: the destructive effects of family pride play a part in many books and films, whether they end happily or sadly.

The other theses fare better. Example B uses a specific quotation and makes a claim that would not apply, for instance, to Hamlet. Example C certainly talks about a specific character in Romeo and Juliet, but it would pass this test better with a specific explanation of what exactly Mercutio's language tells us "about how the play works."

If your thesis is specific enough, it will make a point that focuses on only a small part of the text you are analyzing. You can and should ultimately apply that point to the work as a whole, but a thesis will call attention to specific parts of it. Let's look at those examples again. (This is the last time, I promise.) One way of spotting the problem with example A is to note that a simple plot summary would support its point. That is not true of the others: example B points to the "near-misses" it mentions, and example C will concentrate on Mercutio's speeches.