Note: these recommendations are geared toward researchers in the arts or humanities.
Developing a research topic is an iterative process, even for a short paper. This is a process that emerges in stages, and one which requires critical (but not criticizing) engagement with the evidence, literature, and prior research. The evidence can be an object, an artifact, a historic event, an idea, a theoretical framework, or existing interpretations.
Ultimately, you want to be able to pose a research question that you will then investigate in your paper.
If you are writing a paper for a course, the initial critical ideas and theoretical frameworks may come from your course readings. Pay attention to footnotes and bibliographies in your readings, because they can help you identify other potential sources of information.
As you are thinking about your topic, consider what, if anything, has already been written. If a lot of literature exists on your topic, you will need to narrow your topic down, and decide how to make it interesting for your reader. Regurgitating or synthesizing what has already been said is very unlikely to be exciting both for you and for those who will be reading your wok. If there is little or no literature on your topic, you will need to think how to frame it so as to take advantage of existing theories in the discipline. You may also be able to take advantage of existing scholarship on related topics.
There are two common types of research papers in the arts and humanities: expository and argumentative. In an expository paper you develop an idea or critical "reading" of something, and then support your idea or "reading" with evidence. In an argumentative essay you propose an argument or a framework to engage in a dialog with and to refute an existing interpretation, and provide evidence to support your argument/interpretation, as well as evidence to refute an existing argument/interpretation. For further elaboration on expository and argumentative papers, as well and for examples of both types of essays, check the book titled The Art of Writing About Art, co-authored by Suzanne Hudson and Nancy Noonan-Morrissey, originally published in 2001. Note that particular disciplines in the arts or humanities may have other specialized types of frameworks for research.
Also, remember that a research paper is not "merely an elaborately footnoted presentation of what a dozen scholars have already said about a topic; it is a thoughtful evaluation of the available evidence , and so is, finally, an expression of what the author [i.e., you] thinks the evidence adds up to." [Sylvan Barnet, A Short Guide to Writing About Art (New York: Pearson/Longman, 2005), 238-239.]
If a lot has been written on your topic, perhaps you should consider the following:
If you stick with a broad topic, you run into the danger of over-generalizing or summarizing existing scholarship, both of which have limited value in contemporary arts and humanities research papers. Summarizing is generally useful for providing background information, as well as for literature reviews. However, it should not constitute the bulk of your paper.
If you are interested in something very specific or very new, you may find that little has been written about it. You might even find that the same information gets repeated everywhere, because nothing else is available. Consider this an opportunity for you to do unique research, and think of the following:
Remember to keep track of your sources, regardless of the stage of your research. The USC Libraries have an excellent guide to citation styles and to citation management software.