The information stored in the genome of an organism has long been thought of in computational terms as a kind of codescript for the construction, operation and control of the system in which it is found. However, genomic sequence information can be interpreted as biological instructions and executed as a genetic codescript only by a suitably prepared cell with which the program is in proper registration. We enquire into the character of the evolutionary process that generates physical systems capable of interpreting, in increasingly elaborate ways, the genetic information they contain. The principle of Informed Generation specifies the need for the spontaneous emergence and evolutionary development of self-organizing processes that generate phenotypes from genotypes. The principle of Informed Generation describes a ubiquitous feature of biological systems: without the prior existence of certain components or functionalities, which are required for the production of themselves and others, no configuration of genetic information that accumulated through Natural Selection could ever serve as a codescript for an organism. The operation of Informed Generation is demonstrated in the stepwise evolution of genetic coding and the general distinction between Natural Selection and Informed Generation is illustrated through consideration of gene-replicase-translatase (GRT) system. It is proposed that Informed Generation represents a quite general process of evolutionary self-organization in biological systems whereby essentially irreversible transitions in the systems' dynamics take them to historically contingent, isolated states whose characteristics are determinants of biological specificity. The operation of Informed Generation may have left detectable traces in topographical features of complex intracellular and ecological networks.