Literature citations

by Geoffrey Hart

Literature citations are how we suggest that readers consult previously published work. As a result, they’re common in the Introduction, Methods, and Discussion sections. Citations are provided for several reasons:

To describe previous research and provide context for the present study.
To describe Methods more completely than is appropriate in a paper that uses those methods but that does not focus on methodology.
To provide results or measurements we could not or did not obtain in our study.
To describe supporting and contradictory results that show where our results fit within the existing body of knowledge.

Providing context

The goal of the Introduction is to summarize your study’s context. This begins with a description of a scientific question that must be answered or a real-world problem that must be solved. Because the goal of your paper is to describe your research, it’s not possible to provide full details of the literature. Instead, it’s necessary to summarize. The cited papers will provide the missing details for readers who need them. The Introduction’s literature review should identify problems that have been solved, problems that have not been solved (including contradictions among previous studies), and which unsolved problems you focused on in your study.

Describing methods

Most research fields have standard methods. Citing those methods lets readers who do not already know them read the cited paper and learn the method. Readers who understand the method will know what you did and how you did it, and can use that knowledge to evaluate your choice of methods and any limitations of your results. If you used other methods or developed your own, it’s necessary to explain why. What methodological problem were you trying to solve? If your chosen method has limitations, cite papers that discuss those limitations and state how you accounted for those limitations. If there are multiple ways to obtain a given measurement, explain why you chose one of these methods (e.g., because it works better for your experimental system)—or try to obtain the same data using two or more methods to provide triangulation. (See my article on “Designing effective research” for details.) Here, “experimental system” refers to the organism, process, system, or environment that you studied.

Providing additional data

It’s never possible to measure everything you want to measure, particularly since your results often reveal phenomena or possible mechanisms that you didn’t know existed before your study. When it’s not possible to perform additional experiments to describe these phenomena, you must instead develop a hypothesis about what causes might explain the results. Data cited from previous research may support or contradict that hypothesis and provide suggestions for future research. Even research that used a different experimental system may support your hypothesis—though the greater the difference between the experimental systems, the weaker the support.

It’s also possible that you measured only some components of your experimental system (e.g., a specific protein or species), but previous authors measured different components (e.g., a different protein or species). Taken together, the current and previous results provide a more complete description of the experimental system.

Comparison with previous research

You generally can’t solve all the problems described in the Introduction. In that case, previous research may provide information you could not obtain. Some previous results will support your conclusions. For example, if you studied a species that has not been studied before, but research on other species found similar results, this supports a hypothesis that similar mechanisms exists across multiple species. Contradictions may invalidate your results; for example, this may happen if many studies obtained different results and you did not provide triangulation and replication to prove that your results are correct. Often, contradictions reveal something important: insights into how the experimental systems differed. Sometimes you can cite references that explain the differences in ways that make the contradiction seem less important. For example, contradictory results may be obtained under different experimental conditions (e.g., different climates orsoils) or using different methods. Explaining those differences provides more confidence in your results and can potentially improve understanding of both the present and past studies.

Language of the cited publication

When you write papers for journals that publish in a non-English language (e.g., Japanese), you know that the journal’s readers understand that language. It is therefore appropriate to cite papers published in the journal’s language in addition to papers from the international English literature. It’s particularly appropriate to emphasize papers published in a non-English language when comparable research has not been published in English. However, you must still cite papers from the international literature, since most research is performed in many countries. In contrast, when you write for an English journal, most of the journal’s readers can read and understand English—the journal’s language. But you cannot know what other languages they understand. Thus, even if you must cite non-English papers, it’s important to look for results in the English literature to ensure that readers will be able to read the cited papers. This is particularly true for general principles, such as standardized methods. Where you can’t provide an English citation, you must provide more details of the non-English study so readers can fully understand the cited research without reading the original paper.

The reviewers who will decide whether to accept your paper for publication can only read papers in their own language—or English for an English journal. If you haven’t cited English papers, they may be unable to understand what you did or what previous researchers did. In that case, you must provide an English summary. In many fields, you can guess who will review your paper, particularly if the journal asks you to propose a list of reviewers. Cite relevant papers by these reviewers, since they will expect you to cite their work.

Conclusion

There are many other questions about how to cite papers. Please don’t hesitate to ask me for additional explanation.

For more information on literature citations, see my book Writing for Science Journals.