Why 4(3) codons?

An attempt is made to answer the question posed by the title of this paper. First we show that in primitive self-replicating oligoribotide systems, selection depended from the very start on the existence of 4 kinds of ribotides, forming 2 complementary pairs. Further selection required that the condensation reactions involving the two last positions of the 3'-end of a growing oligoribotide fragment and the first position of the 5'-end of another fragment were catalyzed by randomly synthesized peptides. This established a codon--amino acid concentration correlation and clinched triplet segments as the basis of the translation process. Finally, physical arguments are given to show that the monochirality of the ribotides arose from stereochemical reasons, as firstly described by Wald, but that of the amino acids is the result of natural selection acting during the peptide-assisted stage of oligoribotide growth.