Only about 1/20 of DNA of higher organisms codes for proteins, by means of classical triplet code. The rest of DNA sequences is largely silent, with unclear functions, if any. The triplet code is not the only code (message) carried by the sequences. There are three levels of molecular communication, where the same sequence ‘‘talks’’ to various bimolecules, while having, respectively, three different appearances: DNA, RNA and protein. Since the molecular structures and, hence, sequence specific preferences of these are substantially different, the original DNA sequence has to carry simultaneously three types of sequence patterns (codes, messages), thus, being a composite structure in which one had the same letter (nucleotide) is frequently involved in several overlapping codes of different nature. This multiplicity and overlapping of the codes is a unique feature of the Gnomic, language of genetic sequences. The coexisting codes have to be degenerate in various degrees to allow an optimal and concerted performance of all the encoded functions. There is an obvious conflict between the best possible performance of a given function and necessity to compromise the quality of a given sequence pattern in favor of other patterns. It appears that the major role of various changes in the sequences on their ‘‘ontogenetic’’ way from DNA to RNA to protein, like RNA editing and splicing, or protein post‐translational modifications is to resolve such conflicts. New data are presented strongly indicating that the gene splicing is such a device to resolve the conflict between the code of DNA folding in chromatin and the triplet code for protein synthesis.