Editorial policy

Divorce content on the internet is unusually bad. Not maliciously, mostly: it is written quickly, repeated from other articles, and rarely checked against the law it describes. Statistics circulate for years with no source. National claims get made about rules that differ in all fifty states. Nobody corrects anything, because nobody is tracking what was said.

This page is our answer to a fair question: why should you believe what you read here?

Every legal claim traces to a source

We keep a claims registry: a single internal record of every legal or factual assertion this site makes. One claim, one entry. Each entry carries the source it was verified against, the date it was checked, whether the rule is national or state-specific, and how confident we are.

Nothing on the site is supposed to exist outside that registry. When a page and the registry disagree, the page is wrong and gets fixed.

We verify against primary sources first. Statutes, federal regulations, court opinions, and government agency material outrank everything else. When a claim rests on something weaker, we say so in the writing rather than dressing it up.

We publish our uncertainty. Every registry entry is tagged. Some are verified against primary law. Some rest on a reputable secondary source and are labeled as such, which means we deliberately do not sharpen the wording beyond what that source supports. A few are editorial judgment rather than law, particularly around safety, and those say so too. Where we could not verify something, the page hedges instead of guessing, and the hedge is the honest signal, not a stylistic tic.

We do not publish a statistic we cannot source. If you have read much divorce content you have met the same unsourced percentages repeatedly. When we cannot find where a number came from, we leave it out or describe the finding qualitatively. Where research is genuinely contested, we say that too, rather than picking the tidier side.

The rule that fixes most bad divorce content

Divorce is state law. The single most common error in this field, ours included before we caught it, is stating a rule as though it were national when it varies.

So every claim we hold carries a jurisdiction tag, and our writing standard is that a rule which varies must never be stated as an absolute. When we cite a specific state, we label it as an example rather than letting it stand in for your state. Roughly a third of what we assert is state-specific, which is a good measure of how often a confident national answer elsewhere is simply wrong.

The coach uses the same facts

Divorcery's AI coach draws on the same claims registry the website does. That is deliberate: it means the answers you get in the app cannot drift from what we publish here, and a correction made once reaches both.

We test it. A standing set of questions, built from the claims most often gotten wrong, is re-asked whenever the registry changes or the underlying model does, and the results are recorded. How the coach works covers what it is and is not.

How this gets written

Divorcery is two people, both of whom have been through this. Every legal claim is checked against a primary source before it goes up, rather than against anyone's confident memory of what the rule is. That step is the whole point: text that sounds authoritative is easy to produce, and text that is correct is not.

What you will not find here is content written to fill a keyword. Every page exists because people actually ask that question.

Corrections

If something here is wrong, we want to know, and we will fix it rather than argue.

Email support@divorcery.com with the page and what you believe is incorrect. A source helps but is not required. Corrections to legal substance update the registry as well as the page, so the fix reaches the coach at the same time.

We do not silently rewrite claims that turned out to be wrong. If a correction changes the substance of an answer, the page reflects the corrected position openly rather than pretending the earlier version never existed.