How introductions are built
The pattern is consistent enough across fields that it has a name.
Three moves
The linguist John Swales described research introductions as three moves, a model known as CARS (Create a Research Space). It's a useful check on a draft: can you point to where each move happens?
| Move | What it does | Typical opening |
|---|---|---|
| Establish the territory | Shows the topic matters and what is known | "Recent studies have shown..." |
| Establish the niche | Points to a gap, a conflict or a question | "However, it remains unclear..." |
| Occupy the niche | States what this paper does | "Here we examine..." |
Think of a funnel
Start at the width of your readers' shared knowledge and narrow steadily to your specific question. For a specialist journal that starting point is already narrow, so you can skip the general case for why the field matters.
Each paragraph should be narrower than the one before it. If one widens again, it's in the wrong place.
How long
For a journal article, an introduction is commonly around a tenth of the paper: three to five paragraphs in many fields. Thesis introductions are longer and are usually a chapter of their own.
The full review of the literature belongs elsewhere. An introduction cites what is needed to justify the question.
End with the aim, not a tour
Finish with what the study does and, where your field expects it, the hypothesis or main finding. A paragraph listing what each section contains is conventional in some fields, such as computer science and economics, and unnecessary in most others.