In the early 1990s, the linguist John Swales collected introductions from published research articles and read them as arguments rather than as background. The same sequence kept appearing. The writer first maps a field. A hole in that field comes next, then a claim about the hole.
He called the pattern Create a Research Space, the CARS model. Writing centers still use it to teach how to write an introduction for an academic paper. The objections in a rejection letter map onto those parts: a missing gap often comes back as incremental, a missing or buried claim as an unclear contribution. Long territory with a late gap reads as a review article.
Key takeaways
- Published introductions usually run three parts in order: territory (the field), gap (what is unresolved), and claim (what this paper contributes).
- The gap sentence is what makes the results feel necessary, and it is the part writers omit most often.
- Mensh and Kording’s context-content-conclusion scheme shows the same narrowing inside each paragraph.
- A paragraph that will not take a territory, gap, or claim label, citations that stay dense to the last line, and a claim that does not match the abstract are the usual signs that a part is missing or out of place.
What are the parts of an academic paper introduction?
An academic paper introduction has three parts: territory, gap, and claim. Swales named the same sequence establishing a territory, establishing a niche, and occupying the niche.
The funnel is the shape of the opening: wide at the field, narrow at the contribution. Mensh and Kording, in Ten Simple Rules for Structuring Papers, describe that same narrowing at paragraph scale: context, then content, then a conclusion the next paragraph can pick up.
The pattern is strongest in empirical research articles. Theoretical papers and some humanities articles stretch the territory, but the gap and the claim still arrive, or the paper reads as unfinished. The model was built on published introductions, so it describes openings that survived review. Rejected drafts are not in the data.
How do the three parts work together?
Reviewers already have names for the failure of each part:
| Part | What it does | Typical signals | Reviewer wording when it is missing |
|---|---|---|---|
| Territory | Places the topic and shows where the field stands | Centrality claims, topic generalizations, citations of prior work | “Not a good fit” or “insufficient grounding” |
| Gap | Names what remains unresolved | Contrast markers, limitation statements, questions | “Incremental” or “unclear motivation” |
| Claim | States the contribution and previews the findings | Purpose statements, principal findings, structure announcement | “Overclaims” or “contribution unclear” |
That last column is where desk rejection often starts. Editors screening for fit read the territory and the claim, and a mismatch between them can end the paper before the results are read.
A methods paper on sample-size reporting runs the same sequence. Territory locates the work in the literature on reproducibility and reporting standards. The gap names a specific failure: power calculations omitted or misreported in a defined set of journals after a reporting guideline was introduced. The claim states what this paper did with that hole, for example measuring the rate and testing whether a checklist changes it. If the topic is a protein or a language-model benchmark, the sentences change and the order of the parts stays the same.
What does the territory paragraph do?
It says why the topic matters now, what the field already knows, and which prior work the argument will rely on. Swales treats those as steps inside the first move. They can sit in one paragraph or two, depending on the journal’s length.
The UMass Amherst Writing Center’s CARS guide lists the stock openings, including “In recent years, researchers have become increasingly interested in…”. The phrasing is conventional on purpose, recognizable at a skim across fields.
When territory runs long, it crowds out the gap, and reviewers meet the contribution only after they have already classified the paper as a survey. Those paragraphs also date a manuscript. Citation recency is a long-standing reviewer habit, and a territory built from references a decade old reads as a project that stopped moving before the writing did.
What does the gap sentence do?
It creates a reason for the paper to exist. Without it, the reviewer supplies one, and that invented reason is what they review. The niche can open as a counter-claim to a prior result, an explicit gap, a question, or a continuation of a research tradition. The signals are consistent enough that the UMass guide quotes them as templates: “However, these studies have failed to recognize the …” does most of the work in published papers.
“More research is needed” names no unknown. “Whether X holds in Y remains untested” does.
The wording matters less than the mechanism. A stated gap creates a demand that the results then answer. Pat Thomson, on the patter blog, ties this move to warrant: the rationale for why the project needed doing. A one-sentence warrant is what a skimming reviewer will actually register; what they usually write in the report is “so what”.
What does the claim paragraph do?
It answers the gap. The usual contents are a purpose statement, the principal findings, and a sketch of the article’s shape. Mensh and Kording argue that the final paragraph should also summarize the results compactly, so a reader knows what is coming before committing to the rest of the paper. Editors and reviewers read that summary against the abstract during triage. A contribution worded differently in each place reads as a paper that has not decided what it argues.
What shows that a part is missing?
Paragraph labels are the fastest signal. Each paragraph in an introduction that narrows does one job: territory, gap, or claim. A paragraph that will not take a label is usually summarizing a paper that belongs in the discussion, and two paragraphs in a row with the same label is where the introduction stops narrowing. Where the citations sit shows where the territory ends: they crowd the opening and thin out as the gap and the claim take over. Full density through the last paragraph is a literature review in an introduction’s format.
Contrast markers locate the gap. Words such as however, yet, remains unclear, and little is known either appear or they do not. None of them means the demand for the paper was never stated. One of them buried mid-paragraph means the demand exists but hides from anyone skimming. The last paragraph’s contribution sentence and the abstract’s contribution sentence are read together, and near-synonyms between them are what reviewers treat as overpromising. Dedicated academic tools, Fukuro among them, can flag that wording drift. They cannot decide which wording is the actual contribution.
What goes wrong in introductions?
The history-of-the-field opening spends two pages on territory, then places the gap in the final paragraph, after a reviewer has already filed the paper as a survey. The results may be new. The opening does not show that. The invisible gap is quieter: it summarizes prior studies in a neutral voice and never says what they leave unresolved, so the report comes back as “unclear motivation” even when the summary is accurate.
The buried claim appears once, in the last sentence of the last paragraph, or in an early paragraph and never again. Reviewers quote the first contribution-like sentence they find, which is not always the one the author would pick. Synonym drift is a related failure: the contribution is worded one way in the introduction, another in the abstract, and a third in the cover letter. Each version can be defensible on its own. Together they read as three papers.
Writers skip the gap more often than the other two parts. Naming a hole in other people’s work can feel rude, and summarizing their papers can feel polite. Reviewers read that politeness as missing motivation.
References and useful policies
- Swales JM. Genre Analysis: English in Academic and Research Settings. Cambridge University Press; 1990. See also the UMass Amherst Writing Center CARS guide.
- Mensh B, Kording K. Ten simple rules for structuring papers. PLoS Computational Biology. 2017;13(9):e1005619.
- Thomson P. Create a research space. patter.
- Thomson P. What is a ‘research warrant’?. patter.
- Nature. How to construct a Nature summary paragraph.
- Thomson P. Getting over bad/limited advice: journal article introductions. patter.