The twenty-four claims in full

How the collection was done

Claims were taken as stated, in the words used, without interpretation. A claim counts once however many times it was made — "a company is a place" was said in some form on nine separate occasions and appears here once.

What was collected:

• Anything asserted as a reason for or against the proposal.
• Anything that functioned as a reason, including statements of preference that were offered in support of a position.
• Not questions, not procedural remarks, not restatements of the proposal itself.

The sorting was done afterwards, from the words, by one person. It is not infallible and two of the twenty-four are arguable — both of them are noted below.

The shape of it

Twelve of twenty-four answered in a kind that cannot settle them. Five never answered at all, four of those because they were never said out loud.

The two arguable sortings

Worth flagging, because a record that admits no doubt about itself should not be trusted.

"The commute is part of the job that was accepted." Sorted here as a value: it says what accepting a job commits you to. It could be read as an interest claim about who agreed to bear what, and somebody sorting it that way would not be wrong. Either way it was answered with a study, which cannot settle it under either reading.

"Collaboration tools have replaced what the office used to do." Sorted as empirical, because it is a claim about whether a replacement happened. It is close to being a definition — it depends entirely on a list of what the office did, which nobody has written. Sorted as empirical and recorded as having no evidence either way, which is the same place it would have ended up under the other reading.

What is not in the table

Anything identifying. No names, no job titles beyond "the people setting the rule", no sector, no city, no numbers that would locate the organisation.

Also absent: how many people held each claim. Several of the value claims were held by nearly everybody on one side and appear here once, the same as a claim made by a single person. A count of claims is not a count of support, and reading the left column as a vote would be wrong.

What the table cannot show

It cannot show whether any of the claims are true. Nine empirical claims sit in this record and the table says nothing about which of them hold — only about whether anybody could find out.

That is the limit of this whole method, and the next part of the site is about what follows from it.

How to read the table

Three columns are doing different work and it is worth reading them in a particular order.

The kind column first, on its own. Read down it without looking at anything else. Nine, ten, five — and notice that the second column is the largest. An argument everybody experiences as a dispute about evidence is, by count, mostly a dispute about what matters.

Then the answer column, next to it. This is where the crossings show. Eight studies produced at value claims, four principles produced at empirical ones, and four claims with no answer at all.

Then the notes. Most rows have none. The ones that do are the exceptions — the claim that was answered properly, the cost that was said aloud once, the claim whose kind could not be read.

What the record cannot settle

The obvious objection to a table like this is that a different person sorting the same twenty-four claims would produce a different table.

Partly true, and the size of the difference matters. Two of the twenty-four are genuinely arguable and are flagged. The other twenty-two sort the same way for anybody applying the test, and the headline numbers barely move: shifting both arguable rows to the other column changes nine-ten-five to eight-eleven-five, and changes nothing about the conclusion.

What a different sorter could not change is the count of crossed answers, which does not depend on the kind of the claim being read correctly in the arguable cases — both of those were answered with a study, which cannot settle them under either reading.