#third party analysis

If you've spent any time in discussions of toxic dynamics — whether the antagonist in question is a family member, an ex, a bullying coworker, a bad boss, a nosy neighbour, a "Karen" customer, or an online interlocutor — you've probably encountered the acronym JADE: Justify, Argue, Defend, Explain. The standard advice is well-rehearsed: don't JADE with antagonists. Don't try to justify your choices, don't argue your case, don't defend yourself against bad-faith characterizations, don't explain your reasoning in the hope that clearer articulation will produce understanding. The reasoning behind this advice is solid. Antagonists aren't operating in good faith, so the usual tools of good-faith communication don't work on them. Justifying invites further interrogation. Arguing legitimizes the frame. Defending concedes that there was something to defend against. Explaining gives them more material to twist. The more you JADE, the more rope you hand them.

That much is well-trodden ground. What's discussed less often, and what I want to focus on here, is a particular downstream consequence of JADEing that makes the dynamic even harder to escape: when a third party reads the textual record of an exchange, JADE often looks indistinguishable from escalation.

This is a real problem, and it deserves more attention than it gets.

The grammar of analytical reading

When a neutral observer — a friend you show the screenshots to, a family member you describe the argument to, a coworker reviewing the email thread, an HR rep reading the Slack history, or even an AI chatbot you ask "what do you think of this exchange" — approaches a transcript, they typically read it in what we might call analytical mode. Analytical mode treats every utterance as a tactical move in a debate. It reads for who said what, who responded with what, who reached further, who deployed bigger concepts.

Inside analytical mode, JADE and escalation look identical. Both produce the same textual artifact: someone saying more, reaching further, invoking larger principles, getting more articulate or more impassioned as the exchange continues. The analyst can't see the difference between someone trying harder to be understood and someone trying harder to win. From the outside, both look like "increasing intensity."

But the interior of these moves is completely different. JADE has a specific interior: it's the experience of feeling unheard and assuming that clearer articulation will fix that. Escalation has a different interior: it's wanting to dominate, hurt, or score points. They produce similar text but they come from opposite places. The analytical eye, by design, can't see interiors. It only sees the page.

The composure asymmetry

There's a second, related problem. Antagonists tend to perform composure. They say less. They deploy short jabs, status moves, mockery economically delivered, and exits framed as dismissals ("I'm done with this conversation," "this is hysterical," "wow, okay"). The earnest party, by contrast, says more — because they're trying. They're working to be understood. They reach for additional framings when the first one doesn't land.

On a transcript, "said less" reads as composure. "Said more" reads as agitation. The person performing detachment looks above-it-all. The person actually engaging looks like they lost their cool. The analytical reader rewards the performance of calm over the substance of engagement.

This means that the textual record is structurally biased in favour of antagonists. They know this, often implicitly. It's part of why they perform the way they do — they're banking on the fact that any third-party review will favour the party who said less and looked less invested. "I was just having a laugh, look how worked up they got." The medium itself does work for them.

Why reaching for stronger framings backfires on the page

There's a particularly painful version of this trap. When the earnest party is trying to articulate something about ethics, principle, or proportionality, they often reach for stronger moral language to make their point land. They invoke larger concepts — fairness, harm, freedom, authoritarianism, dignity — because they're trying to gesture at why the antagonist's behaviour is wrong, not just that it's annoying.

But on the page, reaching for stronger moral language reads as disproportion. The analyst plucks out the strongest comparison and grades it as overreach. "You compared X to Y? That's hysterical." The disproportion isn't actually in the engager's view — it's a function of how hard they're working to be heard against someone who refuses to hear. But the textual record can't show that distinction. It just shows the comparison, sitting there, looking immoderate.

This is one of the cruellest mechanisms in the whole dynamic. The harder you work to be understood, the worse you look to anyone reading later.

The closing-move problem

Here's another piece that gets lost in analytical readings: closing moves are nearly invisible on the page.

In a healthy exchange, there's often a moment where one party signals they've gotten what they came for. "Okay, that makes sense." "Alright, fair enough." "That's all I needed to know." These are closing moves — attempts to wrap up the exchange amicably. In a healthy exchange, the other party accepts the close and the conversation ends.

With antagonists, closing moves get refused. The antagonist won't let the exchange end on the engager's terms, because the engager being satisfied breaks the antagonist's script. So the antagonist reopens the conversation, often with mockery or a new tangent. From that point on, every subsequent reply from the engager isn't a continuation of the original argument — it's an attempt to re-close a door that keeps being reopened.

But on the page, "trying to re-close" looks identical to "continuing to argue." The analyst sees turn N of M and counts it as one more round of debate. They don't notice that turn N was supposed to be the last turn, and that everything after it exists only because one party refused to let the exchange end.

This is why exchanges with antagonists often have a strange shape: the engager tries to leave amicably, gets pulled back in, tries to leave again, gets pulled back in again, and finally — sometimes after many rounds — gives up or gets visibly frustrated. The analyst reading later sees only the final frustration and concludes that the engager was the unreasonable one. They miss the entire pattern of attempted exits.

Why all of this matters

The reason the don't-JADE advice is so important isn't just that JADE doesn't work on antagonists in the moment. It's that the textual record of JADEing will be used against you — by the antagonist themselves, by mutual acquaintances, by HR, by family courts, by anyone who later reads the exchange in analytical mode. The transcript is rigged against the person who showed up genuinely. The medium favours performance over engagement, brevity over articulation, dismissal over reaching.

This is a hard thing to internalize, because it runs against the intuition that "if I just explain clearly enough, anyone reading this will see what really happened." They won't. Not in analytical mode. They'll see the surface shape — increasing intensity, stronger comparisons, more words from you and fewer from them — and they'll draw the conclusion that surface shape suggests.

The protection against this isn't better JADEing. It's not JADEing at all. It's recognizing that the moment you start trying to be understood by someone who's committed to misunderstanding you, you're producing a record that will read as your fault no matter how careful you are with your words.

A note for anyone carrying an old exchange

If you've been sitting with a transcript of an old conflict — an email thread, a comment exchange, a chat log — and part of what's kept it raw is the worry that "anyone who read this would side with them," I want to name something directly: that worry is often well-founded, and it's not because you were wrong. It's because the textual record of any exchange with an antagonist is biased against the person who engaged in good faith. A casual analytical reader probably would read it the way you fear. But that's a flaw in casual analytical reading, not a verdict on you.

The reader who sees you as the unreasonable one is missing the interior — what you were doing, what you were responding to, what you were trying to close, what kept being reopened. The interior is where the truth of the exchange lives. The page can't show it. That's the limit of the medium, not a fact about you.