We’re trained, from the first day of school, to finish what we start. A paragraph needs a closing sentence. An essay must tie everything up with a tidy bow. A debate should end with a clear winner. This hunger for closure runs so deep in how we think and write that we often confuse a well-packaged conclusion with a well-reasoned one. But what if the most honest, most productive, and most intellectually alive kind of argument is the one that refuses to end? What if the incomplete argument—the one that trails off, contradicts itself, or leaves a question hanging—isn’t a failure of rigor, but its highest form?

I’ve been chewing on this for years, watching how conversations play out in academic seminars, in comment sections, and over dinner tables. The arguments that stick with me are rarely the airtight ones. They’re the ones with a crack, a seam, a spot where the speaker paused and said, in effect, “I don’t know what comes next.” There’s a peculiar intellectual courage in that pause, and we don’t talk about it nearly enough.

The Tyranny of the Final Draft

We live in a culture that fetishizes the final draft. The published paper, the polished op-ed, the viral thread that lays out a case with bulletproof confidence—these are held up as models of good thinking. But anyone who has ever actually thought deeply about something knows that the final draft is a lie, or at least a carefully staged performance. It erases the detours, the dead ends, the moments of genuine confusion that made the insight possible in the first place.

This performance serves a social function, of course. We want our experts to sound certain. We want our leaders to project clarity. But the cost is a kind of collective intellectual dishonesty. We learn to hide the scaffolding of our thoughts, presenting only the finished edifice. And in doing so, we make it harder for others—and for ourselves—to understand how thinking actually works.

Consider the classic five-paragraph essay, that stalwart of high school English classes. It demands a thesis, three supporting points, and a conclusion that restates the thesis with a sense of triumphant return. The form itself forbids uncertainty. You cannot, in the fifth paragraph, suddenly realize your thesis was wrong and leave the reader with an open question. The structure won’t allow it. And so students learn, early on, that good writing means pretending you never changed your mind.

The Hidden Pedagogy of the Unfinished

But outside the classroom, the most generative thinking often happens in fragments. Notebooks are full of half-formed ideas. Letters between scientists and philosophers are littered with speculation, backtracking, and admissions of ignorance. Darwin’s notebooks, for instance, are a mess of crossed-out hypotheses and marginal queries. They are not arguments in the traditional sense; they are the raw material of argument, and they are far more instructive than the polished prose of On the Origin of Species.

This suggests a strange pedagogical inversion: we teach students to write finished arguments, but we learn the most from unfinished ones. A completed argument is a product; an incomplete one is a process. And process, not product, is where understanding actually grows. When we read a thinker who is still visibly wrestling with a problem, we are invited into the wrestling match ourselves. We don’t just absorb a conclusion; we participate in the struggle to reach one.

The Epistemic Virtue of the Trailing Off

There is a particular rhetorical move I’ve come to admire, one that has no name in the standard handbooks. It’s the moment when a writer or speaker, having built a careful chain of reasoning, suddenly stops and says something like: “But then again…” or “Of course, this all falls apart if…” or simply “I’m not sure.” This is not a rhetorical trick. It’s an epistemic virtue—a willingness to hold one’s own conclusions lightly, to acknowledge the limits of one’s own frame.

We might call this the “trailing off.” It’s the opposite of the rhetorical flourish. Where the flourish aims to sweep the audience along on a wave of conviction, the trailing off invites the audience to pause and reflect. It says: I’ve taken you this far, but I won’t pretend the path continues smoothly ahead. You’ll have to pick your own way from here.

This move is surprisingly rare in public discourse, and its rarity is telling. Pundits are paid to have takes, not to have doubts. The format of cable news and social media punishes the trailing off with a loss of airtime or algorithmic visibility. But in private conversation, among people who trust each other, the trailing off is where the real work gets done. It’s the signal that we’ve moved from performance to exploration.

Incompleteness as Intellectual Hospitality

An incomplete argument is also a form of hospitality. A closed argument says: I have figured this out, and I will now explain it to you. The relationship is hierarchical. An open argument says: I have been thinking about this, and I’ve gotten somewhere, but there’s more to do, and you might be able to help. The relationship is collaborative.

This distinction matters enormously for how knowledge actually advances. The history of science is not a parade of completed arguments; it’s a messy conversation full of partial insights that get picked up, modified, and extended by others. The double helix wasn’t discovered in a single eureka moment; it emerged from a tangle of incomplete models, wrong turns, and competitive collaboration. Watson and Crick’s famous 1953 paper is remarkably tentative in places, openly acknowledging the limits of their data and the possibility of alternative structures. That tentativeness didn’t weaken their contribution; it opened space for others to build on it.

The Aesthetics of the Fragment

There’s also an aesthetic dimension to this. Fragments have a power that wholes often lack. A ruined temple can be more evocative than an intact one. An unfinished symphony can haunt the imagination in ways a completed one never could. The same is true of arguments. An argument that leaves a gap creates a kind of intellectual negative space, and the mind rushes in to fill it.

The Romantic poets understood this well. Coleridge’s “Kubla Khan” is famously a fragment, and its prefatory note claims it was interrupted by a visitor from Porlock. Whether that story is true or a fabrication, the poem’s fragmentary status has become part of its mystique. We don’t just read the poem; we imagine what it might have become. The incompleteness is generative.

In philosophy, too, the fragment has a distinguished pedigree. Nietzsche’s aphoristic style is a deliberate rejection of systematic completeness. Each aphorism is a provocation, not a conclusion. It demands a response. Pascal’s Pensées are scraps of a projected apology for Christianity that was never written. Their power lies precisely in their unfinished quality—they feel like thoughts caught in flight, not thoughts embalmed in a treatise.

The Risk of Romanticizing the Unfinished

But we should be careful not to romanticize incompleteness for its own sake. There’s a difference between an argument that is incomplete because the thinker is honestly grappling with difficulty, and an argument that is incomplete because the thinker is lazy or evasive. The former is a sign of intellectual integrity; the latter is just a sign of bad work.

The key distinction, I think, is whether the incompleteness is acknowledged. An argument that pretends to be complete but is actually full of holes is not an incomplete argument in the sense I’m defending; it’s a failed complete argument. The virtue lies in the explicit recognition of limits, the deliberate choice to leave a question open rather than paper over it with a fake answer.

This requires a certain kind of intellectual maturity. You have to know enough to know what you don’t know. You have to be secure enough in your own competence to admit its boundaries. And you have to trust your audience enough to show them the scaffolding instead of just the finished facade.

Incompleteness in Public Life

What would it look like to bring this ethos into our public discourse? Imagine a political debate where a candidate, when asked a hard question, said: “I don’t have a fully worked-out answer to that. Here’s what I’m thinking so far, and here’s where I’m stuck.” The immediate reaction would probably be scorn. Pundits would call it a gaffe. Opponents would seize on it as weakness. But some portion of the audience—the portion that’s tired of canned talking points—might feel a flash of respect.

We’ve been conditioned to see uncertainty as a liability. But in a complex world, certainty is often the real liability. The problems we face—climate change, inequality, the governance of artificial intelligence—are genuinely hard. No one has a complete answer to them. Pretending otherwise is not leadership; it’s theater. And theater, however satisfying in the moment, doesn’t solve problems.

There are a few public figures who have managed to make incompleteness part of their intellectual brand. The late philosopher Daniel Dennett was known for his willingness to say “I don’t know” in interviews, often followed by a thoughtful “but here’s how I’d start thinking about it.” That combination—humility plus a demonstration of how to think—is far more valuable than a pre-packaged conclusion. It models the process, not just the product.

The Digital Paradox

The internet, for all its flaws, has created new spaces for incomplete arguments. A blog post can be updated, revised, or left deliberately open-ended. A Twitter thread can unfold over days, incorporating feedback and changing direction. A podcast conversation can meander without needing to reach a verdict. These formats are closer to the natural rhythm of thinking than the old print formats ever were.

But the internet also punishes incompleteness. The algorithm rewards the hot take, the definitive thread, the thread that goes viral because it seems to settle something. And the permanence of the digital record makes people hesitant to leave traces of their uncertainty. A half-formed thought posted today can be screenshotted and used against you a decade from now. So we self-censor. We wait until we’re sure—or until we can fake sureness convincingly—before we speak.

This is a loss. The internet could be a vast laboratory of incomplete arguments, a place where ideas are tested and refined in public view. Instead, it often becomes a museum of finished takes, each one sealed in its own vitrine, untouchable and inert.

The Case for the Unfinished Essay

Let me try to practice what I’m preaching. This essay itself is, in a sense, incomplete. I’ve been circling a question—what makes incomplete arguments valuable?—and I’ve offered some partial answers. But I haven’t tied them all together into a neat theory. I haven’t even defined “incomplete argument” with precision. And I’m not going to.

Because the moment I try to define it precisely, I’ll start excluding things that might be interesting. The moment I try to build a systematic defense, I’ll have to paper over the cracks in my own reasoning. Better to leave the cracks visible. Better to let you, the reader, decide which parts are useful and which parts aren’t, and how they might fit together—or not.

This is not an abdication of responsibility. It’s a different kind of responsibility. Instead of taking responsibility for giving you a complete answer, I’m taking responsibility for giving you an honest process. I’m showing you my thinking, not just my conclusions. And I’m trusting you to do something with it.

How to Write an Incomplete Argument Well

If you’re going to write an incomplete argument—and I think more of us should, at least some of the time—there are a few principles worth keeping in mind.

First, do the work. An incomplete argument is not an excuse to skip the hard parts. You should still research, still reason carefully, still anticipate objections. The incompleteness should come from genuine difficulty, not from laziness.

Second, signal your incompleteness clearly. Don’t just stop; tell the reader why you’re stopping. Is it because the data runs out? Because a counterargument you can’t answer has occurred to you? Because the question turns out to be deeper than you thought? The reader deserves to know what kind of incompleteness they’re dealing with.

Third, leave handholds. An incomplete argument should not be a blank wall. It should offer places for the reader to grab on and continue the climb. Point to the next questions. Suggest possible paths. Indicate where you’d go next if you had more time or more insight.

Fourth, be honest about your confidence. Distinguish between the parts you’re fairly sure about and the parts you’re not. Use phrases like “I’m reasonably confident that…” or “This is where I get speculative…” or “I’m genuinely torn here.” This granularity of confidence is far more informative than a blanket assertion or a blanket disclaimer.

The Deeper Question

Underneath all of this is a deeper question about what we want from arguments in the first place. Do we want them to settle things, or do we want them to unsettle things? Do we want them to provide answers, or do we want them to provoke better questions?

The answer, of course, is that we need both. There are times when a definitive conclusion is exactly what’s called for—a mathematical proof, a legal verdict, a medical diagnosis. But there are also times—and they are more frequent than we admit—when the push for closure is premature, when the honest move is to keep the question open.

Our intellectual culture has over-indexed on closure. We’ve built our institutions, our formats, and our habits of mind around the finished argument. It’s time to make more room for the unfinished one. Not as a replacement, but as a complement. Not as a sign of weakness, but as a sign of strength.

I’ll end here, not because I’ve said everything there is to say, but because I’ve said enough to start a conversation. The rest is up to you.

Frequently Asked Questions

Isn’t an incomplete argument just a fancy way of saying you didn’t finish your work?

Not if it’s done deliberately. There’s a difference between abandoning an argument out of laziness and choosing to leave it open because the question genuinely resists closure. The latter requires more intellectual honesty, not less. It means you’ve thought deeply enough to recognize the limits of your own reasoning, and you’re willing to show those limits rather than hide them behind a forced conclusion.

How can an incomplete argument be persuasive? Don’t people want clear answers?

Persuasion isn’t always about providing a clear answer. Sometimes it’s about changing how someone thinks about a question. An incomplete argument can be deeply persuasive in that sense—it can shake someone out of a settled view and open up new lines of inquiry. And in many contexts, people actually find honest uncertainty more trustworthy than overconfident certainty. We’ve all encountered experts who pretend to know everything and end up being wrong; the expert who says “I’m not sure” often earns more respect in the long run.

Doesn’t this approach just lead to endless navel-gazing and no real progress?

It can, if taken to an extreme. But the point isn’t to replace completed arguments entirely; it’s to make space for incompleteness where it’s appropriate. Many of the greatest intellectual advances started with someone saying “I don’t know” and then working through the problem in public view. The incompleteness is a stage in the process, not the end of it. The danger is when we skip that stage entirely and jump straight to a fake completeness that short-circuits real thinking.

How do I know when to leave an argument incomplete versus when to push through to a conclusion?

Ask yourself whether the conclusion you’d reach would be genuine or forced. If you can arrive at a conclusion that honestly reflects the weight of the evidence and your own reasoning, then by all means conclude. But if you find yourself papering over gaps, ignoring strong counterarguments, or pretending to more confidence than you actually have, that’s a sign that incompleteness might be the more honest choice. Also consider your audience: are they looking for a definitive answer, or are they capable of engaging with an open question? The context matters as much as the content.