Life as Interplay of Information and Matter

Any sequence can be transformed into an equivalent one by alphabet change and/or encoding. Thus an information cannot be identified to a single sequence, but should be defined as the equivalence class among sequences with respect to such transformations, an abstract entity. It may be represented by its ‘information message’ defined as the shortest binary sequence in this class, whose length quantitatively measures the information. Any sequence must be borne by a physical medium, so information bridges the abstract and the concrete. Perturbations in the physical world result in random symbol errors, while information may instruct the assembly of physical objects by the agency of the semantics it bears. Information theory tells that errorless communication is possible provided a long and redundant enough error-correcting code is used. Conserving genetic information over the ages needs error-correcting codes making the exact regeneration of genomes possible provided it is performed frequently enough. In engineering, such codes are conveniently defined by mathematical constraints which make the symbols mutually dependent, but constraints of any kind, defining ‘soft codes’, also do the job. The better conservation of old parts of the genome implies moreover that the genomic error-correcting codes are made of nested component codes. This scheme accounts for basic life features: need for successive generations, existence of discrete species with hierarchical taxonomy, trend of evolution towards increased complexity, etc. Transcription of the genes and their translation into polypeptidic chains, becoming proteins when properly folded, are controlled by enzymes, which as proteins are are needed for their own synthesis. A feedback loop results, referred to as ‘semantic’ since it implements the genetic code which tells the meaning of the genes. Once closed, it keeps its structure, thereby conserving the genetic code. The onset of a new semantic feedback loop originates an organic code in Barbieri’s meaning, which itself induces a new soft component in the system of genomic nested error-correcting codes by imposing new constraints to the genome. Due to the feedback, constraints incurred by the proteins affect the genome. Besides the synthesis of proteins, the assembly of living structures, as instructed by the genome, involves interwoven semantic feedback loops. Although closed, they do not prevent the genome lengthening, say by horizontal genetic transfer, which increases both the information quantity, hence its semantic specificity, and the redundancy, hence its error-correction ability. Thus, closing semantic feedback loops does not prevent evolution: new structures can be appended with improved resilience to errors. Regeneration failure is very infrequent and results in a widely different genome which originates a new species if the phenotype it specifies withstands the Darwinian selection. A possible scenario for the origin of semantic feedback loops (hence maybe of life) is proposed.