What is code and does DNA have one?

A coding function maps a set of objects to a set of symbols: O → S , where S is a set of symbols (tokens); O is a set of objects. A decoding function is the inverse and maps the set of symbols back to the set of objects: S → O. Without loss of generality S can be thought of as a set of configurations of a material system. The set of coding function values is determined by the semantic cargo voluntarily attached by the decision making agent to specific configurations. 

E.g. for an N-digit lock, the coding function is a Boolean function returning TRUE for a single arbitrarily chosen sequence of digits and FALSE for any other sequence. Here the choice of the unlocking sequence which is made by the owner of the lock does not depend on physical constraints apart from the length N of the sequence. 

In the case of the genome symbols are codons (triplets of nucleotides) and objects are the amino acid residues of the protein molecules being synthesysed. 

The possibility of using code in practice as well as the practical possibility of detecting it in configurations of a material system depends on the following being enabled: 

• independence of a particular configuration from the physical constraints; 
• transmission of code through a channel in order to record, store and interpret it; 
• interpretation/decoding; 
• stability of the mapping between symbols and objects. 

Coding activity is based on making choices that are not constrained by the laws of nature i.e. inert to them. E.g., the choice of the rules of chess does not depend on the laws of nature even though the realisation of the game of chess is subject to them simply because it takes place in the physical world in space and time. In non-living nature there exists no coding because non-living nature is free from any rules and is only subject to constraints. On the contrary, code is inherently rule-based and is, consequently, unique to artificial systems and biological systems. Therefore I believe that it is highly likely that code generation in biosystems must have involved agency similarly to artificial systems involving code.  

Is there any justification for claims made by some that recording, reading and processing of genetic information in biosystems are wholly determined by the physics and chemistry of the processes/reactions involved?! There is not a single law of nature which under given initial conditions can produce systems generating and processing code. Existing evolutionary models using only chance and necessity causation are inadequate and unrealistic. On the contrary, practice testifies to the fact that code is exclusively a product of purposive agency. Consequently, any system utilising code is an artifact as a whole since it has components responsible for coding/decoding of information.