About
Popular claims, followed back to the source
Last updated .
A claim starts in a study, a survey or a speech. By the time it reaches a headline it has usually been shortened, rounded and detached from the people it was about. Nobody has to be dishonest for that to happen. Each step just repeats the step before it.
Predicatio takes one claim and walks the chain backward. For each step it records the wording, the date, what the writer cited, and what the cited source says when I open it. Each step gets a plain verdict: supported, overstated, distorted, a change in what the words mean, unsourced, circular, or a source I could not open. Then the piece sets out what the strongest available evidence does show, so a trace does not end as a debunk when the underlying question is real.
How a claim gets picked
I look for claims that are widely repeated, that matter to decisions people make, and whose origin few readers have seen. A claim whose trail goes cold is also a result, and I report it as one.
What this site is not
It does not decide who is right in a political argument and it does not guess at anyone's motives. A piece says what a source states and what a later writer said it stated. It does not give medical, psychological or legal advice. Where a topic touches mental health, the piece points to support and avoids detail that could do harm.
Who's writing this
I'm Christopher Ross. I run thisismyurl.com, a web design and strategy practice, and I'm finishing a master of arts in Learning and Technology at Royal Roads University. Predicatio is my own project. It is not client work, nobody sponsors it, and it is not for hire by the people and companies it covers.
Research, drafting and checking here use AI assistance as a normal part of the workflow. Every citation is checked by script against the source, every factual sentence is tied to a passage in a source the script opened, a different model checks each claim against its passage, and separate AI reviewers with different briefs read the draft. Nothing publishes unless every one of those checks passes. A piece that misses the bar slips its date. I audit a sample of published pieces afterward. The standards page shows what each check does.