Translation is a merger, not an invention

The peptidyl transferase centre and the aminoacyl-transfer-RNA system are treated here as two RNA- world artefacts subject to selection independently, in a world with no code and nothing to translate. The first is a generator : a ribozyme that condenses ambient activated amino acids into untemplated peptide. The second is a carrier : a stable, heritable association between a particular amino acid and a particular sequence, written on the acceptor stem and legible without the anticodon. Neither was maintained on account of the other. The genetic code is read as the frozen interface contract of their encounter, recording which charging specificities obtained when the two lineages came to occupy one compartment. The account requires selective, not functional, independence, which makes it indifferent to whether the components are related by descent and therefore compatible with accounts that derive them from one duplicated ancestor. The interface is identified with the CCA terminus, on three independent lines of evidence, and the chemistry of the encounter has been performed: a scaffold derived from the transferase centre, charged acceptor-stem minihelices, and nothing else, forms a peptide bond. Demonstrated RNA-amino acid affnities are shown to concern the complex, late-recruited amino acids rather than the founding set, so that chemistry biases the code without determining it. A test is proposed: the same scaffold, charged with two different amino acids, should yield a dipeptide whose composition tracks the charging.