Crossed answers
Catching yourself doing it
Everything else on this site is easier to apply to other people. Four signals that you are the one crossing, and the repair.
Claims this part is about, and what each was answered with
A company is a place, not a login
Value≠A studyYoung people are owed the experience of working near others
Value≠A studyPeople are less likely to leave if they feel connected to a place
Empirical≠A principleHybrid teams make worse decisions than co-located ones
Empirical≠A principlefour claims · two empirical and two value · four were answered with the wrong kind of answer
Four of the twelve crossed answers on this record.
The sorting is much easier to apply to somebody else's claims than to your own, and a site about it is liable to become a way of diagnosing opponents.
Four signals that you are the one crossing. None requires anybody else to point it out.
One: you reached for the study first
If your first instinct on hearing a claim was to retrieve a piece of evidence, check what kind of claim it was. The retrieval happens faster than the sorting, which means it happens whether or not the sorting would have licensed it.
The habit to build is a half-second pause between hearing and reaching. That is the whole of it, and for anybody who reads research it repays more than any other change on this site.
Two: they restated instead of answering
The most reliable external signal, available in every exchange.
If the person repeats themselves rather than conceding, countering or qualifying, your reply did not reach their claim. They may be being stubborn. More often they have not been answered, and the restatement is the only move left to them.
Three: you would not change your mind either
Ask of your own position what you would ask of theirs: what finding would move me?
If the answer is nothing, then whatever you have been arguing has been a value all along, and every study you produced was decoration. That is worth knowing before somebody else notices it.
Four: you are enjoying it
The least respectable signal and a good one.
A correctly sorted exchange about a genuine disagreement is mostly uncomfortable: you are conceding sortings, naming your own values, and proposing trades. An exchange that feels like winning is usually one where you have answered something that was not asked, because that is the only kind you cannot lose.
The repair, in your own words
If you catch one, the fix is short and it does not require a confession:
Let me take that back — I answered a different thing. What you said was about what work should be, and I gave you a productivity finding. On what work should be, here is what I actually think.
Three clauses: withdraw, name the kind, answer in kind. About fifteen seconds, and it is the only move on this site that reliably changes how the other person argues, because it hands them a model of the thing without requiring them to read anything.
The limit of self-correction
It will not fix a discussion by itself. On this record, one participant began sorting their own claims from week four onwards and the discussion still went to attribution and then to procedure.
What it did produce was the two exchanges in eleven weeks that reached a trade. That is a small return on a habit that costs nothing, and it is the whole of what one person can do without anybody else's agreement.
The fifth signal, which is the best one
One more, and it outperforms the other four: check whether you have ever moved.
Over the whole of this argument, in your organisation, have you changed your position on anything? Conceded a point, accepted a measurement, revised a claim?
If the answer is no — not softened your tone, actually changed your mind about something — then whatever you have been doing has not been argument. It has been advocacy, which is a respectable activity and a different one, and advocacy produces crossed answers by design because its purpose is to advance a position rather than to find out.
Most people in these discussions have not moved on anything. That is worth knowing about oneself before diagnosing it in others.
What moving once does
More than any technique on this site.
A person who has visibly conceded something is believed afterwards in a way that nobody else in the room is. It costs one point, usually a small one, and it buys credibility for everything after it.
On this record one participant conceded, early, that their central claim was a value rather than a finding. They were also the only person whose later contributions were consistently engaged with on their merits rather than discounted as positioning.
The written test
The most reliable version of all five signals requires no self-knowledge at all: write down your own three central claims and hand them to somebody uninvolved.
Ask one question: which of these could evidence settle?
Their answers will differ from yours on at least one, usually the one you are most confident about. That is the whole test, it takes five minutes, and it does not depend on your ability to be honest with yourself in a moment of advocacy — which is the faculty the other four signals rely on, and which nobody has.
Why this is worth doing before the discussion
Because the discovery is cheap in advance and expensive in public.
Finding out that your strongest claim is a value, alone, before a meeting, costs you nothing and lets you restate it well. Finding out during an exchange, when somebody else points it out, costs the claim and the standing that went with it.