The Working Conclusion is a publication about how we decide what to believe, how we disagree, and what the available evidence actually permits us to conclude.
The subjects vary. Politics, power, media, history, persuasion, institutions, and the arguments people have across a kitchen table. What connects them is one problem. Human beings are very good at reaching conclusions and remarkably bad at showing one another how they got there.
Every serious piece here does six things.
The documented record. What can actually be established. Primary documents, original statements, official data, transcripts, direct video.
The relevant research. What studies found, and what they did not establish. Population, method, size, limitations. A study is evidence, not a magic spell.
The strongest counterargument. The most credible competing explanation, not the weakest version I can beat. When it explains part of the evidence better than my account does, I say so.
My inference, labelled as inference. Interpretation is unavoidable. Pretending otherwise only hides it. When I move from what the evidence establishes to what I think it means, you’ll see exactly where that happens.
A working conclusion. I answer the question. Intellectual honesty is not an excuse for permanent indecision.
Revision conditions. What evidence would change the answer. A conclusion that can’t be changed by evidence isn’t a conclusion. It’s a belief being protected from examination.
I have a worldview
I’m not going to promise political neutrality. I have values, experiences, and political judgments, and they don’t disappear because a camera turns on.
My obligation isn’t to become a person without a perspective. It’s to stop perspective from masquerading as evidence. When my worldview materially affects a piece, I disclose it before the evidence rather than after.
The hardest scrutiny goes to claims I want to believe, not only to claims I dislike. That’s when the method matters most.
On bridges
This publication is trying to build one. But a bridge is not the midpoint between a truth and a lie.
It runs from a claim to the evidence.
I’ll listen seriously to competing arguments. I won’t manufacture equivalence the evidence doesn’t support. And I won’t confuse open-mindedness with refusing to reach a conclusion.
Corrections
I’ll get things wrong. The test isn’t whether that happens. The test is what happens next.
Corrections here are visible, dated, specific, and attached to the original work. Each one states what was originally said, what was wrong, what the better evidence shows, and whether the conclusion changes. Every correction is collected in the Evidence Ledger.
A correction made because better evidence appeared is not a failure of the system. It’s the system working.
Who I am
Byron Williams. I founded The Working Conclusion LLC, which is also making Mirrorturn, a structured turn-based process for two people to work through one specific disagreement. This publication is about how arguments work. Mirrorturn is how two people actually run one.
What’s free
The weekly piece, the podcast, the sources, the strongest counterargument, the working conclusion, and every correction are free and always will be. You should never have to pay to inspect the factual basis of a public argument.
Paid membership adds depth. Expanded source annotations, evidence that didn’t make the final cut, research notes, and a monthly discussion. It doesn’t buy access to the reasoning.
The best answer the evidence supports. For now.
