Writing an Introduction

The Introduction has one job: convince the reader that a specific problem matters, that something about it is unknown, and that this paper does something about it. Everything else is decoration. If a reader finishes the section still unsure what your paper asks or why they should care, the section has failed no matter how much background you packed in.

This page complements scientific writing, which places the Introduction inside the IMRaD structure. Here the focus is the shape of the section and the moves that make it work.

The funnel#

An Introduction is a funnel. It starts broad with the field and why it matters, narrows to the specific gap in what we know, and ends at the single question this paper answers. By the time the reader reaches the bottom, they should believe three things: the area matters, something specific is missing, and this paper supplies it.

The funnel maps onto three questions asked in order.

  1. What is the topic, and why does it matter?
  2. What is the problem, the gap in current knowledge?
  3. What did this paper do about it?

In the simplest case each question is one paragraph. Three or four paragraphs is the usual length. Five or six is a soft ceiling, and going past it almost always means background has crept in that belongs in the Methods or the Discussion.

The however moment#

The hinge between the topic and the gap is the part that carries the section. The reader moves along with the broad context, then hits a sentence that makes them stop and think that something here is not known. That pause is the reason the paper exists.

You do not need the word “however,” but the gap has to be unambiguous and bounded. “Little is known about X” is the classic weak version. It is usually false, always unconvincing, and tells the reader nothing about what your paper will measure. Name the specific thing that has not been measured, tested, or resolved.

Ending on the question#

Close the funnel with what the paper does, framed as filling the gap you just named. One sentence stating the question or objective is the minimum. A second sentence previewing the approach or the main finding helps the reader anchor before they reach the Methods.

For modelling and hypothesis-driven work, the “to learn X, we did Y” pattern is the safe default. Pose the question, then say in a sentence or two how you went after it. Do not drift into full Methods prose. A line of orientation is the limit.

The opening sentence#

The first sentence is your one chance to make a reader keep going. Skip the empty openers. “Scientists have long known” and “There is a long-running debate about” fail the read-aloud test and waste the position.

Stronger openings do real work. A surprising quantitative fact, a reframing that shifts how the reader sees the topic, or a tension the paper will resolve all earn the reader’s attention. Coordinate this sentence with the opening of the abstract so the two do not repeat.

Common failures#

Three failures account for most weak Introductions.

The literature review in disguise. If your sentences make the researcher the subject (“Smith found X, Jones found Y”), you are cataloguing what is known instead of arguing toward a gap. Rewrite so the finding is the subject, with the citation at the end (“X is the case”). The Introduction describes the hole in the wall, not the whole wall.

The disconnected gap. The gap in paragraph two has to match what the paper actually delivers. A gap that sounds important but does not connect to your aim leaves the reader confused about why you did what you did.

The buried question. If the reader cannot point to a sentence that says what this paper does, add one. It belongs at the end of the section, stated plainly.

A worked skeleton#

Paragraph 1, the topic. Opening sentence that grabs attention (a surprising fact, a reframing, or a tension). Why the area matters, shown rather than asserted.

Paragraph 2, the problem. Narrow from the topic to a specific aspect. Name the concrete, bounded gap (the however moment), and say why closing it matters for the bigger question.

Paragraph 3, what you did. One sentence stating what this paper does, framed as filling the gap. A brief preview of the approach, and optionally the main finding or hypothesis.

A worked infectious-disease example#

Symptom-based isolation is the default response to a new respiratory pathogen, and it assumes that people become infectious around the time they feel unwell. When transmission runs ahead of symptoms, that assumption breaks and the strategy leaks. For pathogen X the timing of infectiousness relative to symptom onset has not been quantified, so it is unknown how much transmission escapes isolation that waits for symptoms. Here we estimate the pre-symptomatic fraction of transmission for pathogen X from contact-tracing pairs and use it to predict when symptom-based isolation fails to hold the reproduction number below one.

Four sentences carry the funnel. The topic and why it matters, the however moment where symptoms and infectiousness come apart, the bounded gap, and the question the paper answers with a one-line preview of the approach.