Read these only after you have written your own three sentences. Two of the paragraphs are good, one is not. The question is not which side is right — the lesson does not settle that — but what a paragraph has to do to count as an argument at all.

Model A — the machinery worked

The machinery worked, because the honesty finding was exactly the kind of claim Popper wants: it forbade a result — no difference between signing at the top and at the bottom — and so outsiders could check it, and did. Falsifiability never promised speed; it promised that a public claim stays open to refutation by anyone, and the anonymous researchers and three bloggers with no institutional standing were that “anyone”. The price I pay is the decade: for ten years a fabricated result functioned as knowledge in insurance forms and government trials, and I have to admit that “correctable in principle” protected nobody who relied on it in the meantime.

What it does: names the side; uses one tool (falsifiability) and says what it does and does not promise; names the price and admits it is a debt.

Model B — the machinery broke down

The machinery broke down, because the people it relied on were not doing the checking. Ariely and Gino received a credibility excess — in Fricker’s terms not an injustice to them, but the reason reviewers, editors and citing authors took the numbers on trust — while the outsiders who did the refuting worked unpaid, partly anonymously, and were sued for $25 million. The price I pay is that the fraud was caught, retracted and punished; I owe an account of what the machinery working would have looked like, and “caught before publication” may be a standard no science has ever met.

What it does: the same three moves, on the other side, with the other tool. Notice that it uses credibility excess exactly — it does not call the famous authors victims of testimonial injustice.

Model C — a paragraph that does not count

I think the machinery failed because Gino was a victim of testimonial injustice — people stopped believing her because she is a woman. Also, the study was never really falsifiable, since it was fabricated, so it was pseudo-science all along. In the end the truth came out, which shows that science works.

What goes wrong:

  • “Testimonial injustice” is asserted without the test. Gino was disbelieved because of evidence — the fabricated numbers — not because of identity prejudice; step 2 of Fricker’s test fails, and the paragraph never runs it.
  • “Never falsifiable, since it was fabricated” confuses the claim with the evidence. The claim forbade a result and was falsifiable; the fabrication made the evidence worthless.
  • It takes both sides at once: “the machinery failed” and “science works”. There is no position left to defend.
  • No price. Nothing in the paragraph costs the writer anything.

The checklist

A paragraph counts as an argument when it:

  1. names a side;
  2. gives a reason that uses one of the lesson’s tools correctly — falsifiability, or Fricker’s credibility terms;
  3. names the price the side pays, in the writer’s own words;
  4. answers, or at least states, the strongest objection from the other side.

Model C fails 2, 3 and 4. Go back to your own paragraph and mark it 1–4.