The most difficult part of any research paper is picking a topic. There is no sure-fire way to do it, and there is no systematic way to approach it. This section will outline some of the basic considerations in thinking about how to pose a historical question which will stimulate your own interest as well as provide a workable approach to a research paper.
What is a historical question?Edit
Every discipline has its own sorts of questions which it tries to answer. A question in astronomy, for example, might be "Why does the sun sometimes have spots on it?" This is a question which stimulated a number of different research programs, using entirely different methods and assumptions, for hundreds of years. It is a question which can be answered using the tools of the discipline and the methodological approaches. A non-astronomical question might be "Is the sun morally good?" It might be a question answerable to a theologian or a philosopher, but an astronomer, using astronomer's tools (be they telescopes or mathematical formulae), would have no way to even begin answering the question.
Similarly, there are questions which fit within the study of history and many which do not. "Was Richard Nixon a good man?" would not be a historical question, and neither would "Was the bombing of Hiroshima and Nagasaki evil?" These questions, while quite valid questions to ask in general, are not generally thought answerable by the historian using only the tools of history.
Many good historical questions are about causes and changes. "Why did the United States bomb Hiroshima and Nagasaki?" is a perfectly good historical question, one on which many books have been written. "Why did the Roman Empire fall?" is another such question. "How did the experience of African Americans change after the Civil Rights movement?" would be another. "How did the SAT become the standard test for American students?" is another. The question need not be stated explicitly in the work itself, but it should be underlying the overall investigation.
There are also many questions which can be quite broader, along the lines of, "What were the experiences of slaves in the Southern United States?" These questions, without specific temporal delimiters, are likely too large to be answered in one research paper — entire books could be written on just this issue. However large questions can be made smaller by the use of temporal or thematic demarcation. For example, by changing our original question to "What were the experiences of slaves in the Southern United States in the years just before the Civil War?" puts a "cap" on how comprehensive our approach must be, and allows us to easily compartmentalize the past in a way suitable to writing a paper.
Historical questions can be very broad or very narrow in scope and it does not necessarily affect how important or interesting the topic may be. What is usually important about whether a topic will be historically "interesting" is whether it "speaks" to a particular larger question or issue.
Searching the literatureEdit
Starting without any topic at all can be difficult, even for the seasoned student. The first step towards finding an area of research is to narrow down the basic range of topics to an existing body of literature. For example, if you were wanting to write a paper related in some way to the discovery of the Americas by Europeans, a good way to start searching for topics would be to find a survey book which would give an overview of the entire historical framework of the time period and topics. A reference librarian should be able to help you find such a thing, as should any expert in the field (do not be afraid to e-mail him if his addresses are posted online).
One easy way to find out which books are currently considered "required reading" for a given topic is to find an academic book or article which is reasonably similar in its own topic, and to look at its first three or so footnotes. Introductions to academic works usually contain one large footnote early on which gives a brief listing of all of the books which any general reader of the paper is expected to be familiar with.
When reading over a survey, keep an eye out for interesting issues or questions you might have. Looking at the footnotes in such works will also give you an idea of the sorts of sources which are available for these sorts of works (there will be more on sources in the next section).
Your topic might change half way into your research, or even while writing up the final paper — this is a natural occurrence and reflects the fact that you have been thinking it over for a longer period of time. Even if you aren't completely sure what your final "question" will be, don't be afraid to start doing the background secondary source research once you have a general idea of what you want your work to be about. History is not an exact science, and many important and interesting things will surface out of sheer serendipity as your work continues
Next section: Finding sources