The synthetases have several active sites that enable them to:
(1) recognize a specific amino acid,
(2) recognize a specific corresponding tRNA(with a specific anticodon),
(3) react the amino acid with ATP (adenosine triphosphate) to form an AMP (adenosine monophosphate) derivative, and then, finally,
(4) link the specific tRNA molecule in question to its corresponding amino acid. Current research suggests that the synthetases recognize particular three-dimensional or chemical features (such as methylated bases) of the tRNA molecule. In virtue of the specificity of the features they must recognize, individual synthetases have highly distinctive shapes that
derive from specifically arranged amino-acid sequences. In other words, the synthetases are themselves marvels of specificity.37
The accuracy of translation depends directly on the specificity of associations between tRNAs and tRNA-binding proteins called aminoacyl-tRNA synthetases (aaRSs). In general, to each type of encoded amino acid corresponds multiple tRNA isoacceptors and a unique aaRS that covalently attaches (aminoacylates or “charges”) that type of amino acid to the ends of those tRNAs at their acceptor stems in an ATP-dependent two-step reaction for later delivery to the A-site of the ribosome for codon-directed insertion of amino acids into growing peptide chains. The tRNAs and aaRSs associated to the same amino acid type are called cognate pairs 1
The original evolutionary stages of the aaRS–tRNA network remain relatively obscure.
That is a grotesque understatement. There is no reasonable explanation whatsoever of how this amazing molecular machinery and assembly factory could have emerged. Why the author mentions evolution is misleading, in face of the fact that this whole process had to emerge prior to DNA replication, and as such, evolution could not have been in play yet.