Open question · part of Truths made by agreement

When a lab result is mangled into a bad headline, the result itself is unharmed, and someone can run the experiment again. Now take another kind of truth: a twenty-dollar bill buys groceries. Mangle that one and there is nothing to re-run, because the truth was the agreement.

The book’s tools were built for the first. The open question was what the chain from source to audience does to the second kind, the “intersubjective” third of Harari’s three realities.

Its working answer: these are real truths, generated by agreement, and passing them on is taking part in them. They cannot be wrong, though they can split or go hollow (the constitutive regimes). Capture means steering whose reading counts, under cover of reification. With no outside referee, a failing agreement gets renegotiated, abandoned or forked, as when Ethereum split into two ledgers in 2016. A captured chain distorts a fact. It decides an agreement.

The book marks this provisionally resolved. Still open: where the line between the two kinds falls (scientific consensus is the hard case), whether three regimes is one too many, and how the account squares with earlier work it cites but has not yet argued with, such as Hacking’s looping effects.

Status: provisionally resolved.

Read the full argument

How it connects


← Reification · Truths made by agreement (6 of 6)

Explore all the ideas · Read the book in order · Open the interactive map