Journal articles — Is it okay to break down the introduction in subsections?

Question

I am writing an article for publication in a journal, and I couldn't find the answer here on Stack Exchange or in the journal guidelines. The field in question is electrical engineering.

As I'm writing my introduction, I would like to tackle different parts of the problematic in a very clear way. I'm thinking about something in the likes of (not necessarily in order):

• Formalizing some definitions and variables that will be used;
• Context of the problem;
• Brief review of models used in the literature (to justify my model of choice);
• Review of parametrization methods for models similar to mine;
• Review of papers tackling the same problem (and therefore the "holes" that my paper will fill);
• Objectives of the article and a brief explanation of the structure (e.g. Section II will show...).

I think that if I write all these points a single section, it will be a huge block of text (even with proper use of paragraphs and writing), and some readers may not pay attention to some important points. So I thought about breaking down my introduction in subsections, as it seems a logical solution to the problem.

I saw some articles doing this, but the majority of journal articles I have read on my area don't do this. Do you guys think it is an acceptable solution to use subsections on an introduction? Can it impact positively or negatively a review process?

Answer 1

Reserve the introduction for

• A very brief background necessary for understanding the next three points
• Summarizing your research question and findings
• Explain how your contribution relates to the existing literature
• Outline the remainder of the paper.

These are four to six paragraphs.

All of your literature review should go into a separate section, after the introduction.

Answer 2

When writing the paper, think about how others will read it. Here is how I typically read engineering papers:

Typically I'm looking for papers on topic X. I usually have a long list (10 - 100) of papers that may or may not be relevant. I can't read them all in depth, I have to "triage" them to determine which ones to spend time on. So first I read the abstract. If that sounds interesting, then I skim the introduction and conclusion. If that sounds interesting, then I glance through the figures. If I'm still interested, THEN I start reading text in depth.

I don't know if this is how others read engineering papers, but I suspect many would (at least, I know for sure that none of my colleagues magically have more spare time on their hands than I do)

With that in mind:

• Structured papers with well organized sections, subsections, etc: good, makes it easily skimmable
• Long introduction: bad, I won't read it all anyway unless I've already decided that your paper is relevant based on process above.
• Suggest that your first figure or two should serve as a "graphical abstract". Something where a quick glance at the figure will give a sense for what the paper is about, even without reading any text.

Answer 3

I'm going to suggest that you don't do so much in an introduction. I suspect that reviewers will say the same. Most of what you have listed should be sections of the main part of the paper, not the introduction. If the intro is so long that it gets boring, then readers may not get to the more important stuff.

I recognize that you are trying to be "logical", but that can be a mistake in an introduction. What you want to do is tell the reader why they should read the whole paper, not present most of the paper itself. But your sections seem to include nearly everything except the results and conclusions.

The key in the intro is to say what problem is being attacked, whether the methodology is standard or innovative, and what we can conclude in general from the research. Put the rest of it in sections of the paper itself.

What you seem to be wanting to do is write half or so of an extended abstract in the introduction. I'd suggest not doing that. Capture the reader's interest in a few sentences. If the various sections of the paper itself are properly named and set off, then there is little need even to write what amounts to a table of contents in the intro. Let the paper's structure do that. Point to things that are unique/innovative/interesting in the intro, but not a general outline. The good will get lost in the ordinary if you try to do too much.

---

Personal note: I once wrote a lesson for students that was so perfectly logical, complete, and elegant that no one understood it at all. It was beautiful, but useless for my purpose - and theirs. Logic and precision are often needed (this was Statistics) but understanding is also required or the logic is distracting and gives little insight.

---

However, for something sent to a journal as a first draft you can structure it pretty much as you like. Very few papers ever get published in their original form. If no one objects to your structure, then it is probably fine, though you should also consider the ultimate reader, not just the reviewers, who may have more background than the typical reader.

Answer 4

You certainly do not want six introductory sections --- that is excessive and it will bore your reader to tears. It should be possible to include all the things you have set out here in two sections: (1) an initial introduction without formal definitions/models that gives a non-formal introduction to the the problem and its context, the relevant literature, and what you will do; and (2) a section formalising the problem and expanding on the details of the models in the literature and how your model departs from these. If you find your explanations so long that you feel you need to break these into subsections, that is a sign that you are being insufficiently parsimoneous in your explanation. You should be able to cover all the points in your post in two sections, and it should generally be possible to cover this without so much text as to require breaking into subsections.

1. Introduction

Here is a general problem and its context

Here is some literature about the problem

Here is what my paper will do, with a brief justification

2. Defining the problem

Here are formal definitions of variables, models, etc.

Here is more detailed discussion of models in the literature that are similar to mine

Here is how my model will differ from those in the literature, and more details on the justification ...