Scientific writing 3.0: A reader and writer's guide - Jean-Luc Lebrun, Justin Lebrun 2021

Structure Q&A
Headings-subheadings: the skeleton of your paper
Paper structure and purpose

Q: The structure is like an outline. Shouldn’t we start a paper by creating its outline first?

A: Some writers use the structure as a framework for writing. They create the structure in bullet point form or in outline form, and then expand the points. This method has value. It gives focus to a paper. If the story flows well at the structure level, it will probably flow well at the detailed level also. The writer may still change the structure later, but it will mostly be to refine the headings or to create more subheadings, not to totally restructure the flow of the paper. Actually, as a side comment, when asked if they know that Microsoft Word includes an outlining program, most scientists say that they are not aware of such a capability! It is right there at your fingertips, an item under the view menu. It comes with its own set of tools. Play with it. I discovered outliners on the Apple II (a program called “More”). Life has never been the same for me since then. Even the first edition of this book started its life as an 11 page outline!

Q: To keep my headings short, can I use acronyms?

A: Using acronyms in headings or subheadings, unless they are wellknown, is troublesome for the majority of readers. Obscure acronyms prevent readers from making a story out of the subheadings. But this is true also of highly specific expert words. Write your headings and subheadings for non-experts. And if you absolutely need to use an expert word in a subheading, define it right away in the first sentence that follows that subheading. Doing so respects the principle of just-in-time background.

Q: What makes a good subheading?

A: A good subheading is an informative noun phrase whose keywords appear with high frequency in the paragraphs that follow it. It fits neatly and logically in between other subheadings. And it is closely related to the contribution, often by including words from the title and the abstract. A bad subheading contains acronyms, synonyms of the title keywords, or specific instances of title keywords (for example the title contains ’memory,’ and the subheading contains ’register,’ a specific keyword not even used in the abstract). The computer science expert knows the connection between memory and register, but the non-expert with a knowledge gap cannot make the connection.

Q: How many paragraphs come under a subheading?

A: That will depend on the length of your paragraphs, and on the flow of your paper. For example, if you describe a succession of steps, but each step is covered in one paragraph of two to four sentences, it may make more sense to group or collapse several steps under a more comprehensive subheading name. Let the specificity of the keywords in the subheading guide you. If the subheading is so specific that its keywords do not even appear in the abstract, reconsider its usefulness or its wording.

Q: My journal dictates the labeling of the subheadings! Can I change them?

A: To enable the reader to reproduce your results, some journals (chemistry or life science journals come to mind) demand that you follow their imposed structure. If a journal dictates the number and types of subheadings, you have no choice but to comply.

Q: I wrote a letter, not a paper. My letter does not have headings, should I include some?

A: Not all papers have an explicit structure. When the paper is short, a letter for example, there is no need to add a structure: the structure is implicit. The “Introduction” heading is absent, but the first paragraph of the letter introduces. The “conclusions” heading is absent, but the last paragraph concludes. However, if you want to benefit from the tests and the metric in this chapter, I recommend that you artificially create and write down such a structure based on the contents of your paragraphs, even though it won’t be part of your letter.

Q: My structure has standard headings (intro, results, ...). Should I include more subheadings?

A: If your paper has enough pages, you should at least add two subheadings (or more) under the heading that contains the bulk of your contribution. It facilitates reading.

Q: I have read the structure of my friend’s paper, and found it very difficult to connect with the title of his paper. What is wrong: the title, or the structure?

A: It could be both. If you were not able to retell the story of the title just by browsing the headings and subheadings of your friend’s paper, you may trace the problem back to five main sources: (1) synonyms (title and structure use different keywords meaning the same thing), (2) Acronyms (defined elsewhere), (3) uninformative headings/ subheadings (hollow markers such as experiment), (4) logical gaps, or (5) unintelligible expert keywords.

Q: Does every single word in the structure need to be taken from the title or abstract?

A: Unless these words are generic words known by all, including well known acronyms, the words in your top-level subheadings need to be taken from the title and abstract. At deeper subheading levels (for example 1.1.2), you can choose any word you like because readers are unlikely to read these subheadings during their first discovery of the paper.

Q: My paper is a review. Do the principles detailed here also apply to this kind of paper?

A: A review paper is quite different, but principle 4 (the story line) should still be respected.