I’ve been thinking about endings lately, prompted by this advice from Kit de Waal about how to conclude flash fiction. “Don’t be tricksy,” she warns. “Don’t bring a rabbit out of the hat at the last moment.” This prompted me to think about the times I’ve read papers, books and dissertations where something goes wrong at the end.

Academic writers often struggle with conclusions in ways that mirror fiction writers. We often know how to begin. And if we don’t, there’s lots of advice on introductions, including mine. And we understand how to build an argument, marshal evidence, engage with theory. Lots of advice about this too. But when it comes to endings, we can stumble. We either say too much or too little. We introduce new ideas we haven’t prepared readers for. We repeat what we’ve already said. Or we fade and die, just as de Waal says.

Let’s start with the rabbit-out-of-the-hat problem. We’ve probably all seen a paper that proceeds carefully, methodically, building an argument. The analysis is solid. The engagement with literature is thorough. And then, in the final paragraph or two, the writer suddenly pivots to something entirely new. A limitation becomes the main point. A tangential policy implication takes centre stage. Or a theoretical framework that hasn’t been mentioned since the literature review suddenly resurfaces as the key interpretive lens.

This feels tricksy because it is. The rabbit-out-of-the-hat disrupts the implicit contract between writer and reader. Now let’s pivot to you writing a paper, not reading someone else’s. Throughout your paper, you’ve been guiding your reader to follow a particular thread, to think alongside you in a certain way. Your conclusion needs to honour that journey. If you’re going to make a significant claim or introduce a crucial theoretical perspective, it needs to be woven through your work, planted in advance, as de Waal puts it. Your reader should arrive at your conclusion thinking “yes, this makes sense, this follows, this feels inevitable” rather than “wait, where did that come from?”

Now to the second problem, starting strong. de Waal’s advises that strong beginnings require equally strong endings that resonate with those openings. In academic writing, we can think about this quite mechanically. The conclusion should “return to” the introduction, should “answer the research questions posed.” (Gah, I’ve said this lot.) But resonance is different from a mechanical return. It’s about creating an echo, a sense of completion that feels organic rather than forced.

Think about your introduction. What question animated it? What tension, problem or puzzle drew you into this work? Your conclusion shouldn’t simply restate that you’ve answered the question. Instead, it should show your reader how the journey through your paper has changed their understandings of the original problem. The ending should feel like it emerges naturally from everything that came before, while simultaneously casting new light on where you began. It’s the difference between a conclusion that says “I told you I would examine X, and I did” and one that says “remember when we started with X? Now look at how differently we can understand it.”

Perhaps the hardest part of de Waal’s advice for academic writers relates to leaving readers with something to do. Academic culture trains us to value comprehensiveness, to spelling things out, to making sure we’ve covered every base and addressed every possible interpretation. We fear that if we don’t explain everything very explicitly, we’ll be criticised for gaps or omissions. And so our conclusions often become exhaustive inventories of findings, implications, limitations, and future directions.

But there’s power in restraint. There’s intellectual generosity in trusting your reader to think alongside you rather than for you. This doesn’t mean being vague or incomplete, you’ve still done the rigorous analytical work, you’ve still made your argument clear. But it means recognising that the best academic writing opens up space for readers to extend your thinking, to see connections you didn’t make explicit, to take your analysis further. I’m not suggesting here that academic conclusions should be cryptic or deliberately obscure. The “don’t force it” and “don’t spell it out” guidance needs to be calibrated for academic genres where clarity and precision matter enormously. But we can over-explain, particularly in conclusions. We worry so much about making sure readers understand our contribution that we end up diminishing it by labouring every single point to the nth degree. But a conclusion can be an invitation to take the work further. As de Wal says, create “a sense that the world you have created goes on.”( I’m thinking about how to exemplify this in another post.)

Next. The “don’t fade and die” warning is perhaps the most common problem that academic writers face. It’s one I often experience myself. Academic conclusions often lose energy in their final paragraphs. After making substantive points, they trail off into increasingly generic statements about limitations or future directions. We’re tired, yes? We’ve got this far, can’t we just stop now? So the final sentences feel obligatory rather than purposeful, as if the writer ran out of steam or couldn’t figure out how to end. This is the academic equivalent of the fiction writer who doesn’t know how to stop, who keeps adding one more paragraph, one more scene, until the story just peters out.

Strong endings in academic work need to maintain energy right through to that final sentence. This doesn’t mean saving your biggest claim for last, that’s the rabbit-out-of-the-hat problem. But it does mean thinking carefully about where to place different elements of your conclusion. If you’re going to discuss limitations, they probably shouldn’t be the final words your reader encounters. If you’re going to point to future research, that gesture should feel generative rather than apologetic. Your last paragraph, your last sentence even, should leave readers with something substantial to hold onto.

And a final key point. de Waal’s advice goes directly to the relationship between writer and reader. Don’t be tricksy. Don’t force it. Leave them work to do. These are instructions for how to respect your reader’s intelligence and engagement while still guiding them toward understanding. They assume a reader who is actively thinking, who will notice if you plant something early that pays off later, who can handle some ambiguity and complexity, who will keep thinking after they finish reading.

Academic writing culture sometimes trains us to assume less capable readers, or at least to write defensively, as if we must anticipate and forestall every possible misunderstanding or criticism. But the best academic prose, like the best fiction, trusts its readers. It gives them a clear path through complex ideas without reducing that complexity. It builds toward conclusions that feel both surprising and inevitable. It knows when to explain and when to suggest. It ends with purpose and power rather than trailing off into disclaimers.

So the next time you’re writing a conclusion, you might revisit de Waal’s flash fiction advice. Have you planted the key elements of your ending throughout your paper? Does your conclusion resonate with your beginning in ways that show genuine development rather than simple restatement? Are you giving your readers the satisfaction of completion while also leaving them space to extend your thinking? And does your ending have the same energy and purpose as your beginning, or have you let it fade and die?

Your conclusion is your last chance to speak to your reader. Make it count. But don’t be tricksy. Let it emerge from the world you’ve built, let it honour the journey you’ve taken your reader on, and trust that they’ll carry your ideas forward into their own thinking and work. That’s when you know you’ve ended well.

Photo by Markus Spiske on Unsplash