How we know things

A page-level bibliography is not provenance. It tells you where somebody looked; it cannot tell you which source supports which sentence, and when a source changes, nobody can work out which sentences just became wrong.

So each stage is written down separately, and each one is checkable:

1. Source — a publication, with an authority profile and the evidence domains it is and is not competent for.
2. Source observation — what that source actually says, with the assumptions it was making attached.
3. Claim — one single assertion, kept with its history, resting on named observations.
4. Reconciliation — where disagreement is resolved, or explicitly preserved as disagreement.
5. Guidance — what we tell you, carrying its reasoning.

A source is not “reliable”. It is reliable about something. A manufacturer is the best source in the world on its own product’s specification and its own manual, and it is not a source on comparative superiority, on neutral market rankings, or on food safety. Those competences are recorded per source and enforced at every citation.

See what each source may and may not be cited for.

A general food-safety rule may only rest on statutory, official or academic sources. Manufacturer and blog evidence cannot establish one, however confident it sounds.

The exception that proves the rule: a manufacturer is authoritative for a safety instruction about its own appliance — empty the grease tray, do not use this model indoors. That instruction is kept to that appliance, which is how both things stay true at once: the maker is the authority on its own product and not on food safety in general.

Every safety rule carries a jurisdiction, an authority, a last-verified date and a review date. Nothing sits on the site indefinitely without somebody agreeing to look at it again.

Barbecue writing is full of confident cultural assertion that traces back to nothing. Repeating it about somebody else’s food culture is a real harm dressed as enthusiasm.

So there is nowhere in BBQHQ for an unqualified cultural assertion to sit. Every such claim carries a qualifier — documented in, commonly associated with, regional variation, not universal, disputed origin, uncertain history — and these four things are refused outright in our own writing:

• Calling one version of a dish the authentic one

  Asserts a single true version of a living tradition. Record it as documented in a named source instead.

• Saying that everybody in a country always or never does something

  A universal claim about a population. Record it as commonly associated, or as not universal, instead.

• Naming a single inventor or birthplace for a dish

  Origin claims need to be marked as disputed, or to rest on strong historical provenance.

• Presenting one regional variant as the tradition itself

  Presents one variant as the tradition. Record it as a regional variation instead.

The visible consequence is that our Traditions section grows slowly: 10 traditions are published, and each one waited for a source that describes the practice rather than a confident summary of it. That is the policy working, not failing.

Where we hold no evidence, the page says so and the question is tracked. There is a difference between something we have established, something no source states, something the sources disagree about, and something that does not apply here — and the page tells you which it is. A blank space looks like an oversight nobody got round to; saying “we do not know” looks like what it is, which is a decision.

The open questions are published for the same reason.

BBQHQ carries no advertising and no affiliate links today. When it does, safety pages, troubleshooting pages that escalate to a safety instruction, and the evidence pages will be excluded — not as a preference but as a rule the site enforces on itself, because a commercial placement beside a warning gives a reader a reason to doubt the warning.