Attributed Is Not Published

We have written four times this month about ways a regulatory claim can be wrong. A claim nobody published. A claim a real page contradicts. A claim that was true when it was written and is still being served. A negative claim with no scope, which cannot be checked because there is nothing to open.

This week produced a fifth, and it is the hardest of the five to catch, because every check we have published passes it.

The sentence

On Thursday evening a local station in East Texas carried a story about search warrants at a chain of smoke shops. By Friday morning the metro dailies had it. The reported figures are large and we are going to set them aside for a moment.

Travelling with that coverage, in more than one outlet, is a sentence about the law:

> The only form of THC allowed in the state right now is .3% delta-9 THC on a dry-weight basis.

Run that sentence through the checks.

Is it attributed? Yes. The outlets credit a press release from the Texas Department of Public Safety. That is a named source, and it is not a blog or an aggregator — it is a state agency with authority over criminal investigation.

Is it fabricated? Almost certainly not. Independent outlets carry it consistently, along with matching detail about agencies and quantities. They plainly have a document in front of them.

Is it stale? No. It is hours old.

Is it an unscoped negative? Not in form. It is a positive statement about what is permitted.

So it passes every test we have written down. And we could not open it.

Where we looked, and when

We read the DPS newsroom on 25 September 2026, at two paths — the news index and the press-release index. Neither carries anything about smoke shops, this chain, the counties involved, search warrants, or THC seizures. The newest item on either index is dated 21 September. We ran a search restricted to the agency's own domain and surfaced nothing on point. We guessed at a release slug and got a 404.

Now the scope, because the scope is the whole discipline and we have been caught on this before.

That is a statement about two indexes and one search, on one date. It is not a statement that no release exists. It is not a suggestion that anyone invented anything. Agencies routinely send material to press lists without posting it to a public newsroom, and there is nothing improper about that. We are not alleging misconduct, and if the release is posted somewhere we did not look, this paragraph is the part of the article that is wrong.

The finding is narrower than misconduct and more useful than an accusation: the most widely quoted statement of Texas THC law this week has no original a retailer can open. If you sell consumable hemp in this state and you want to read the actual wording of the claim being made about the category your inventory sits in, you cannot get to it. You can only read a paraphrase of it.

And with a paraphrase, the question is always what the paraphrase dropped.

What this one dropped

One word. The rule does not bind delta-9 THC.

It binds total delta-9 THC, which is not a number a laboratory measures directly. It is a computed value:

> total delta-9 THC = delta-9 THC + (0.877 × THCA)

The threshold is 0.3% or less on a dry weight basis, inclusive, with an allowance where the accredited laboratory's measurement of uncertainty produces a range that includes 0.3% or less.

Read the quoted sentence again with that in hand. ".3% delta-9 THC on a dry-weight basis" is not a loose restatement of that threshold. It names a different quantity. The word it leaves out is the word that pulls THCA into the calculation, and THCA is the entire reason the computation exists.

The practical consequence is not subtle. A retailer who reads that sentence, picks up a certificate of analysis, finds a delta-9 figure comfortably under 0.3% and stops has not checked the number the rule binds. They have checked one of its two inputs. The certificate may not even carry the other one.

We are not going to re-derive the arithmetic here, because it is already written up and nothing about it has changed. It is at gethempdash.com/learn/total-delta-9-thc-how-it-is-computed, and the reason THCA reaches Texas rules by two separate routes is at gethempdash.com/learn/thca-two-texas-rules.

What is new today is not the arithmetic. It is that the sentence circulating most widely this week states the threshold in a form that cannot be checked against a certificate, and the document it came from is not somewhere you can go and read.

Asked twice, and the answers disagreed

We asked the question an operator would actually type, in two independent phrasings, and recorded what the answers said rather than which pages they listed.

The first phrasing came back settled and confident. It named a single permitted form of THC, at 0.3 percent or less delta-9 by dry weight, and called high-THCA flower, vapes, concentrates and delta-8 products "no longer compliant." Note the same dropped word, and note two assertions about what is and is not permitted that we will come back to.

The second phrasing came back the opposite way. It said the legal status "is currently very unclear." It reported that police in one city described possession of smokable hemp products as lawful while police in another suggested it would be a felony, and that no one could say definitively which. And then, a few lines on, it reproduced the agency characterisation as settled framing anyway.

Yesterday we recorded a pair of phrasings where the control came back clean, and said that a clean control is what makes a split pair mean something. Today there is no clean control. The pair splits, one half calls the matter settled, the other half says nobody knows, and neither of them says which provision it is talking about.

That last part matters more than the disagreement. "The only form of THC allowed" is a sentence about a threshold. "Is possession of smokable hemp lawful" is a question about different provisions entirely. An answer that does not say which one it means will sound like it has answered both.

The surface that points the other way

For the record, and quoted rather than characterised.

The Department of State Health Services publishes the Consumable Hemp Program page. We read it in full again on 25 September 2026 — the sixth consecutive morning we have done so. Under its item recording that on 24 June 2022 the Texas Supreme Court upheld the statute prohibiting "the processing or manufacturing of a consumable hemp product for smoking", the page continues:

> At the same time, the lower court's injunction remains, which allows the distribution and retail sale of consumable hemp products for smoking.

We are not offering that as a refutation of anything above, and it is important to be exact about why not. That sentence is about the processing-and-manufacturing prohibition in the Health and Safety Code. The characterisation travelling with this week's coverage is about the threshold and about delta-8. Those are different provisions, and a sentence about one is not evidence about the other. We draw no conclusion in either direction, we have read no underlying court order, we hold no cause number, and we make no statement about enforcement posture or about whether any particular product may be offered for sale.

What the two surfaces together do establish is narrower and worth saying plainly: two Texas state agencies publish material on this subject, and a retailer reading either one alone would come away with a firmer sense of settledness than the record supports.

One more thing about that page, noted for the sixth time: it carries no date stamp of its own. No "last updated", no "reviewed on". The only date attached to our reading of it is the date we wrote down ourselves.

About the enforcement action

We said we would come back to the figures, and here is all we are going to do with them.

An enforcement event is not an enforcement posture. Our record for the threshold holds the rule text as verified and the enforcement posture as open, and it holds that open in both directions regardless of what any individual case appears to show. Nothing in this week's reporting changes it, and nothing in this article rests on it.

Two specifics are worth stating because they are in the reporting and are easy to skim past. No arrests had been made at the time of reporting, and no charges were named. And the seizure inventory as reported is dominated by material that is not consumable hemp at all — a large quantity of psilocybin mushrooms, and marijuana. No hemp regulator appears anywhere in the coverage.

We have written this lesson once already, about a warrant executed in San Antonio in August that read for a full day as a hemp-enforcement story and turned out, on the affidavits, to be a months-long narcotics investigation with no hemp programme in it. That piece is at gethempdash.com/learn/a-raid-is-not-a-rule-change and we are not re-arguing it. The point of raising the shape again is only this: day-one framing of a raid is frequently not what the case is about, and here there is no charging instrument published at all, so there is not even a document to correct the framing with later.

The practical version

Ask what the paraphrase dropped. When the only available statement of a rule is a summary in a news story, the useful question is not whether the source is reputable. It is which words are missing. Here the missing word was "total", and it was the word that made the number checkable.

Check the number the rule binds, not the number the sentence mentions. For the Texas threshold that means total delta-9 THC, computed from delta-9 and THCA, not the delta-9 line on its own. A certificate that does not report the THCA figure has not given you enough to compute it.

Distinguish attributed from published. A named institution in a sentence is not the same as a document you can open. Both are better than an anonymous claim, and only one of them lets you check the wording. If you are writing the sentence down for your own staff, write down which of the two you had.

Say which provision you mean. Most of the confusion in this week's answers is not disagreement about a rule. It is two different rules being discussed in sentences that never name either.

And keep a negative claim scoped, including ours. We looked in two places on one date and did not find a release. That is what we found. It is not the same as the release not existing, and the difference is the entire reason this paragraph exists.

The rules we track, each carrying the date we read the surface, are at gethempdash.com/texas-rules.