A paper contributes knowledge
The first job of a paper is to contribute knowledge. That sounds simple. But it immediately raises a harder question — what is knowledge? This varies by research field, by tradition, and by the kind of question being asked.
None of these answers is more “real” than the others. They’re different responses to the same underlying question: based on our research, what kind of claim can we make? In other words, how does our research contribute to how we see or understand the world?
So, to write a paper, it helps to know what type of knowledge you’re contributing — and that type of knowledge usually maps onto a particular shape of question:
| Field example | Question shape | Type of knowledge |
|---|---|---|
| Experimental biology | Does X affect Y? | A causal claim, tested empirically |
| Clinical medicine | Does X work in practice? | Practical / applied evidence |
| Mathematics | Is X true? | Proof |
| Theory-building research | How should we think about X? | A framework or theory |
| Descriptive research | What is X? | Description or characterisation |
Once you know the shape of the question you’re answering, you know, roughly, what kind of claim your paper needs to build toward — which is a big part of what “knowing what your paper is supposed to do” actually means.
A paper convinces the reader
The second job of a paper is to convince reviewers that your claims are valid. Unless you’re working on a mathematical proof, chances are your research doesn’t reveal something true. Only something that’s likely true. So you need to show why your claims are likely true.
This means laying out the facts for your audience in a way that lets them follow your reasoning easily. This is largely why we use the IMRaD format — it’s familiar, so your audience already knows what information to expect and where.
Same format, different expectations
The tricky thing with academic papers is that most of them follow the same format: Introduction, Methods, Results, Discussion. Yet which section carries the most weight — and how each one should be shaped — differs widely depending on the type of knowledge your paper is contributing.
Take two examples:
- An applied ecology paper, answering “does X work in practice?”, usually needs its Discussion section to do most of the heavy lifting — because the contribution is in what the findings mean for real-world decisions, not just in the data itself.
- A methods-driven paper, answering “how should we measure/do X?”, usually needs its Methods section to carry the weight — because the contribution is the method, and the reader needs to be convinced it’s sound and reproducible.
Notice that this connects directly back to the table above: the type of knowledge you’re contributing doesn’t just shape what you argue, it shapes where in the paper your argument actually lives. So while two papers might both use the IMRaD format, their internal structure — what’s long, what’s short, what’s doing the persuasive work — can look completely different from field to field.
What this means for writing your paper
This doesn’t mean you need to immediately dive deep into figuring out the type of research question you’re answering or knowledge claim you’re contributing. However, it can be quite helpful to use this framework if you get stuck later on, to ask yourself: what am I trying to do, exactly? So this is more an analytical tool to be aware of, than something you need to get right immediately.
On getting writing advice
Another reason this framework can be helpful is because researchers are often not aware they may be making different knowledge claims — when they’re working in interdisciplinary teams, advising a student working in a field different from their own, or working on a methods project that’s genuinely different from a data-driven paper.
In all of these cases, and others, writing advice doesn’t always carry over as straightforwardly as we’d like to believe. So if you find that specific advice on how to structure your paper isn’t working for you, this may very well be the culprit: you’re working on a different knowledge contribution than whatever the advisor is assuming.