A research paper isn’t a book report in which you summarize and synthesize the analyses that other people have done. In a research paper, you need to be doing your own close readings and then situating those readings in an on-going critical conversation. Go to the texts, read them, interpret them, analyze them, build an argument about them, write the paper that presents and demonstrates that argument, and only then go the scholarship of others.
The point of writing an academic paper, even in an undergraduate classroom setting, is not to be right but to advance a conversation. In other words, the point of academic writing, at least the kind of academic writing I’m asking you to do, is not for your ideas to be correct vis-à-vis the published scholarship on an issue. The point is for you to bring something new to the conversation, however miniscule that something might be (a new question, an unexpected insight, new information, a qualification, a complication, an original interpretation, etc.)
As a rule, therefore, your research and study of scholarship should be conducted relatively late in the process of writing a paper. Doing so will ensure that your discussion occurs on your own terms and in a manner that is uniquely yours. If you do all your research and then try to articulate your interpretation in a paper, you'll often find yourself limited to agreeing and disagreeing with stated opinions, instead of contributing an original interpretation.
In the humanities, as you probably know, there is often a distinction made between primary and secondary sources. A primary source is an object, document, or piece of information that comes directly from the period or topic under consideration. Examples of primary sources would be ancient Aztec calendars, the Magna Carta (1215), William Shakespeare’s Hamlet (1599), the United States Constitution (1787), Anne Frank’s diary (1947), the 2000 census for Saline County in Kansas, or Barack Obama’s address to the Democratic National Convention (2004). A secondary source is a piece of writing, written by someone with some degree of authority on an issue, that offers an interpretation of the objects, documents, and pieces of information that we class as “primary sources.” Examples of secondary sources would include an anthropologist’s book about ancient Aztec astrology, a historian’s article about the origins of modern civil rights in medieval England, a literary critic’s reading of purgatory as a theme in Hamlet, a politician’s claim that our second-amendment rights are in danger, a psychologist’s speculations on the adolescent mind under national socialism, a newspaper article in the Salina Journal lamenting the city’s decreased population, or a public intellectual’s interview on CNN about the rise of Obama’s political star-power.
In short, a primary source is something that needs to be interpreted, and a secondary source is an interpretation. Or, in the terms of my “Aphorisms on the Elements of Academic Argument,” primary sources are textual and historical evidence, and secondary sources are citational evidence.
The distinction between primary and secondary sources is less common in the social sciences than it is in the humanities, but it is still important to recognize. In a social science context, primary sources would be the statistics and examples we use for our interpretations – that is, information for both quantitative (statistics) and qualitative (examples) analysis – and secondary sources would be the already existing scholarship that interprets those statistics and examples for us.
It is true that sometimes we take our basic information (statistics and examples) from scholarly books and articles – that is, from secondary sources. This is not, as a rule, a bad thing. The problem with doing so, however, is that information from secondary sources comes to us prepackaged with interpretation. You as an analyst don’t have the opportunity to examine the information in its pure form, from the ground up. Often, what happens when you take your information from sources that already interpret it for you is that you are reduced to simply agreeing or disagreeing with the claims that are put forth; at the very least, you are likely to conduct your interpretation on someone else’s terms, as opposed to framing and carrying out an interpretation on your own terms in the ways that makes the most sense to you.
To be sure, it is imperative to your research project that you study and address the publications of scholars, but a research project that begins with published scholarship is doomed to be derivative as opposed to original. By this point in your career as a writer (I am addressing these thoughts to college students), a paper that presents an original interpretation, even if that interpretation is not yet fully developed, is infinitely better than a paper that simply reports the interpretations of others and agrees or disagrees with them.
This is not to say that you can totally ignore the published scholarship on an issue; to do so would be fatal to a research project. There will come a point in your project when you’ll need to refer to that scholarship in order to develop your understanding of the issue and to situate your ideas within an ongoing academic conversation. Before you do so, however, you must generate some original insight, perspective, analysis, or argument that might add something to the conversation. At the inception of the research process, therefore, you should willfully ignore the published scholarship and go straight to the basic information (statistics and examples) in question.
Thus, gather up your primary sources, analyze them, formulate your argument, even write an entire paper, and only then look into the scholarship that might help you develop your ideas and situate your interpretation among the ideas of others. If you do so, you may find that the scholarship radically changes (notionally improves) your ideas, and that you must rewrite your entire paper; that is the nature of academic writing. What will hopefully also happen, however, is that some insight, analysis, or argument that you had looking at the information on your own terms is not represented in the scholarship, which will allow you to make a contribution to that conversation.
If, after doing all of your research, you arrive at the opinion that a certain scholar is correct, then we don’t need you to write a paper to tell us so. What we need you to do is to identify any gaps or misconceptions in a scholarly conversation, or to extend the scope and implications of the position that you’re endorsing in some way. If you find yourself writing a paper essentially saying that someone else’s interpretation is right then, sorry, we don’t need you.