What if your proof isn't enough for them?

"Positionality": Agree on the standard and how far it holds

They’re looking at the same numbers

Let’s see if this sounds familiar: It’s the end of the month, and two people are sitting in front of the same budget. Same account, same numbers, same screen.

One of them looks it over and says, “We’re fine. We covered all our expenses this month.” The other looks at the exact same figures and says, “That’s not fine—there’s nothing left over. If anything unexpected happens, we’re in trouble.”

If you’ve ever shared a bank account (or a project budget, or a P&L), you know how this goes from here. It escalates. Somebody pulls up the spreadsheet to prove the point.

But notice: they are not actually fighting about the numbers. They can’t be. They’re looking at the same numbers. What they’re fighting about is what those numbers would have to show to count as “fine.”

And that’s why the spreadsheet won’t settle it. You could hand each of them every spreadsheet in the world, document down to the penny exactly where they stand, and it still might not move either one of them. Because, as I’ve said before, some arguments are not actually about the evidence at all.

This one is about which evidence counts, and for what.

That’s a trickier kind of disagreement to spot, because it hides underneath the part everyone’s arguing about. It’s one of a few I’ve been working through in this series, the times when everything about your idea, or your case, or your argument is right, and the other person still won’t come along.

Last time, the thing hiding underneath was whether the thing was even real. This time, our two budgeters agree it’s real, and they’re still stuck, because they don’t agree on how they’d know whether they’re okay.

There’s a name for the ground you’re standing on when you take a position like that, your positionality, and a name for this particular layer of it: the layer of how you know what you know. (Philosophers call it epistemology. It won’t be on the quiz. Honestly, you don’t even have to remember the word. You just have to notice that this is a different fight than the one about what’s real.)

Here’s the good news, and the whole reason this one is worth slowing down for:

The way out is not more proof, or more evidence.
It’s agreeing, out loud, on what evidence you’d both accept, and on how far it has to hold.

That’s the whole idea, really. But like most simple-sounding things, it has a few moving parts. Let me show you.

More proof isn’t a standard

So what do we usually do when we hit a wall like this? Reach for more evidence.

It’s a natural instinct. If they don’t agree, we think they don’t have enough information yet, so we provide more of it. Better data. A cleaner chart. One more example. Each side shows up with its own stack of proof, each certain that enough of it will finally settle the thing.

But notice what our two budgeters could do with all that proof: nothing.

More spreadsheets don’t help when the disagreement is about what the spreadsheet is even supposed to show. What’s actually missing isn’t more evidence. It’s a standard: the thing you measure the evidence against.

A standard is what lets two people reach the same conclusion without having to take each other’s word for it.

That’s the entire job of a standard. Even better, a shared standard is what makes evidence portable between two people, so that “we’re fine” means the same thing to both of them, and points to the same conclusion for both of them. And note: that standard isn’t the evidence itself. It’s the thing you hold the evidence up against.

Without a shared standard, you can each bring all the proof you want. Your proof, my proof, their proof, our proof. You may may never agree on what the proof says, or what it means, because you’re each measuring it against a different standard—one you never named out loud.

Same standard, still stuck

Now, I’d love to tell you that finding a shared standard fixes everything.

It doesn’t, at least not on its own.

But it’s worth seeing exactly where it falls short, because this is the part we miss too often:

Picture a founder, fresh off a week of customer calls, absolutely lit up. “Customers love this,” she says. “I talked to ten of them, and every one of them said it’s exactly what they need.” To her, that settles it. That’s demand. The evidence is in.

Her cofounder is quieter about it. He says, maybe out loud, maybe half under his breath, “Okay, but who were those ten? Were they people we already had a relationship with? Because if they were, that may not be an accurate reflection of demand.”

Here’s what’s so interesting about that exchange: the two of them don’t actually disagree about the standard. They both accept the same one. Talking to customers is a fair way to read demand; customer conversations count. On that, they’re aligned.

What they disagree about is how far that evidence reaches. She’s treating ten enthusiastic conversations as proof of demand, everywhere, full stop. He’s saying those ten conversations tell you something real, but only here, only about this particular, friendly corner of the world, and only so far.

Same standard. Different reach.

So the disagreement that’s left isn’t about the standard at all anymore. It’s about how far the finding is being asked to travel.

How far a finding reaches

And here’s why that second fight is so predictable: it’s built into the nature of evidence itself.

Any finding only holds as far as the conditions that produced it.

That’s true of ten customer calls, and it’s true of a double-blind study. Whatever you learned, you learned it under some set of conditions, and strictly speaking, that’s the only place you’ve earned the right to claim it. It’s one of those can’t-get-a-lily-from-a-rose-bush things. The rose bush is real and the rose is lovely, but it was never going to hand you a lily.

I know that can sound like bad news, as though nothing is ever really settled. But stay with it for a second, because it’s actually an enormous relief.

It means you never have to defend the impossible.

You don’t have to say, “This is the standard, and it is true in every case, for every person, for all time.” Nobody can win that argument, which is exactly why nobody should have to. You only have to agree on something smaller and far more honest: this is how we’re going to know, in this situation, for now.

So when you’re trying to get somewhere with someone who doesn’t share your view of what’s knowable, the goal was never a standard that holds forever. The goal is a standard with a stated reach: a shared way of knowing, plus a clear limit to how far it goes.

A standard with a stated reach

So let’s go back to our two people and their budget, because this is where it gets practical.

Their whole fight was about one word: “fine.” So instead of arguing about whether they are fine, they can go looking, together, for what “fine” would actually have to mean.

Is “fine” covering all the bills this month?
Is it covering all the bills and being able to absorb any surprise, always?
Or something in between: bills covered, plus a cushion, so that at the end of every month there’s a set amount still sitting there?

Say they land on that last one. Now “fine” isn’t a feeling either of them is defending. It’s a standard they built together, and it comes with a stated reach: it counts here, for this household, this month, as long as certain things stay true. As long as we both have our jobs. As long as we’re saving for the trip. Until the credit card’s paid off. When the conditions change, they’ll revisit it, on purpose, out loud.

Here’s how you can do the same thing, no matter the topic:.

1. Say out loud what would convince you

Start by naming the standard you’re already using, because you are using one, whether you’ve said it or not. One partner’s is “the bills are paid on time.” The other’s is “we can take a hit and be okay.” Neither is wrong. But until they’re actually spoken, they can’t be reconciled.

Most disagreements about evidence are really disagreements about standards nobody said out loud.

So say yours. Then ask the other person for theirs.

2. Pick one standard, together

Then, together, choose one. This is the part people brace against, so let me be clear: picking one standard usually doesn’t mean choosing one of the two you each walked in with. More often it’s a third one you build between you, and you keep working at it until you find the one that works for both of you.

This isn’t winning, and it isn’t compromise. It’s finding the single test you’ll both agree tells you whether the thing worked. And it doesn’t have to be the standard for everything, everywhere. That’s the pressure valve, the safety release, if you will: it only has to hold for this.

3. Say how far it reaches

Then put the fence around it, out loud. “We’re agreeing on this standard as long as these conditions hold.” Which lets you say the one thing that actually gets people to sign on: we’re agreeing on this, here, for now, not on everything, everywhere, forever. That second version is nearly impossible to get anyone to accept. The first one is often surprisingly easy.

4. Check that it actually held

Last, don’t assume the agreement. Confirm it. This is the step everyone skips, and the one that saves you.

Ask, out loud: “How comfortable are you with that, really? Is there any part of it that still doesn’t sit right?”

And mean it when you ask, because inviting the disagreement now is the only way to keep it from surfacing later, bigger, at the worst possible time.

But doesn’t picking the standard rig the answer?

By now you might be having one of two reactions.

The first: “Hang on—if the two of them just pick the standard, haven’t they rigged the outcome? Chosen in advance what will count as knowing, so that of course they’ll ‘know’ whatever they set out to find? Isn’t that deciding the answer instead of discovering it?”

Or maybe the opposite one: “This will never work! Some people are just too far apart. Different worlds, different rules for what counts. Why even try?”

A standard with a stated reach answers both, and it’s the same answer.

Agreeing what would count, for this, for now, isn’t deciding what’s true. It’s agreeing on how you’ll check.

You haven’t rigged anything. You’ve made your test visible and agreed to be held to it, which is the opposite of silently deciding the answer in your own head. And as for “too far apart to ever agree,” that’s almost never true all the way down. You can nearly always find one narrow claim, at one stated reach, that the two of you can both test.

Total agreement was never the bar. Enough to take the next step together is.

And to be clear, I’m not telling you this is the correct way to know things. I just told you what would convince me, and how far I’d let that go. If you were sitting here with me, we could work out a shared version between us. (In fact, let’s: tell me where you’d set the standard for this question, and how far you’d let it reach.)

Agree how you’d know, then how far it holds

So what does all of this come down to, the next time you’re the one who’s stuck?

Remember that you don’t break a disagreement about what’s knowable by out-proving the other person, or by dragging them onto your side of the line. You break it by finding the place, wherever it turns out to be, where your two views actually overlap. The place where you can both say what would count as knowing, and how far that knowing has to hold. Once you have that, however small it is, you have something to build on.

This is the second of three places these hidden gaps tend to open, and they stack. Underneath everything is the question of what’s even real, what’s there to talk about at all. That’s ontology, the layer I wrote about last time. Sitting on top of it is the one we’ve been in today: once we agree something’s there, how do we know it, how do we prove it to ourselves and to each other? That’s epistemology. And there’s one more layer above this one, the trickiest of all, because it isn’t about what’s real or how we know it, but about whether it even matters. That’s value, what philosophers call axiology, and it’s where this series goes next.

For now, though, here’s what I’d love you to sit with.

Think about an argument you’re stuck in right now. Are you really disagreeing about the evidence? Or about what would settle the question in the first place?

Tell me what you land on. I always want to know.

Until next time, Tamsen

Go Deeper

If you want the thinking behind the “some people are just too far apart to bother” objection, read Thomas Kuhn’s The Structure of Scientific Revolutions (University of Chicago Press; originally 1962), the origin of incommensurability, the idea that rival frameworks can share no common measure. What’s less often remembered is that Kuhn kept revising it: he insisted incommensurability never meant the two sides couldn’t be compared at all, and in later work narrowed it to local, taxonomic mismatches, specific terms that don’t translate cleanly between classificatory schemes, rather than whole worlds sealed off from each other. Alexander Bird’s “Thomas Kuhn“ entry in the Stanford Encyclopedia of Philosophy (2022) traces that evolution.

If you want the thinking behind the way through, read David L. Morgan’s “Paradigms Lost and Pragmatism Regained“ (Journal of Mixed Methods Research, 2007). Morgan makes the pragmatist case that you don’t have to settle whose framework is correct to do useful shared work, which is the stance this piece models: agree on what would count, for this, for now, and then check it.

If you want the thinking behind why “known here, for now” isn’t a cop-out, read John Dewey on warranted assertibility, his name for treating knowledge as something earned and warranted through inquiry, held as far as the inquiry supports it, rather than a fixed correspondence to reality settled once and for all. Dewey develops it in Logic: The Theory of Inquiry (1938); his earlier How We Think (1910) is a more accessible entry to his model of inquiry, though it predates the term.