INTERACTIONS.

In the world of Arts, the artist has a peculiar possibility: he can create something and assign a sense only when the work is finished. Jean Arp (1886-1966) said, about his sculptures: << everyone of these bodies has a definite meaning, but it's only when I feel not to change anything more, I decide what all means and I give it a name >>. Now I'm interested to value some of the implications and consequences inferable from this sentence (forgetting the possible provocative content).
First of all it appears clear the operating possibility to distinctly separate the practical realization (that is the definition of a personal syntax) from the attribution of sense. Moreover it's immediate to think at a multiplicity of meanings (all equally permissible), and therefore to a choice.
With Duchamp's ready-made we face a very different case. With them Duchamp talked about the non-necessarity of the definition of a personal language, demonstrating that it was still possible to make meaningful artworks: the artwork is built on the meaning, which is expressed through the ability of the artist to invite to reflect. Provocation and reflection. The reflecting activity becomes inevitable as the only fundamental element as of art and as life.
In both cases attribution of meaning is postponed. Such possibility is not present in common language, nor in computer languages. However in computer languages and in formal systems - and in Arts, too - it's possible to mantain syntax separate from semantics, and to redefine semantics without altering the syntax. No man could do the same thing with his natural language, because (1) we are inclined to confuse nouns with their meanings, and (2) syntax and semantics of words are in our heads too much rigidly connected. Consequently there's one strong link between artistic languages and computer languages.

Now, let's apply to the world of Art a schematization that has its origin in computer science (N.B.: the reasoning that follows is only formal).

We suppose to consider the world V of artist X, in the sense of private psychological world of an individual (i.e. culture and knowledge of X, with its fears, emotions, etc.) as it's seen from the outside, from who observes, through Language. One can consider V like Kandinsky "inner world" too. So we make correspond V to the world M of machine T, and an artwork Oi to a program Pi written in a high level language A (provided with portability features). Pi is interpreted by a suitable interpreter program (or it is compiled).
So the correspondances are:
 
 

In the same  way  we  can  think about  an   interpreter   automaton (human or not) who analyzes an artwork O1, explains it in a comprehensible way, and defines a world V1 as output (with V1 subset of V, if all works well). Analyzing O2,O3,…,On we'll obtain V2,V3,…,Vn, and through their union, an approximation of V : V'=V1UV2 … UVn. Formally, supposing to have built V', one can think that whichever entity (human or automatic) operating on V' could construct a set of works {O1',O2',…,On'} in which X can recognize itself. In this construction it's necessary, for obvious reasons, to prescind from any technical or manual ability. However, the artworks built this way are not injured: it's sufficient to remember the observations of Duchamp.
Now let consider the operating environment of a computer system; we have an external entity (a programmer) who writes a program in a language A: it's up to the interpreter/compiler program to translate the phrases of the A language in sentences comprehensible to the machine. Let's look at the world M (that is the equivalent of V): if the observation of V makes it possible for us to enter in contact with the sweetness, the obsessions, the culture (etc.) of X, what could we obtain from the observation of M? It's not sensible to attribute feelings to a machine: nevertheless, it's allowable to attribute poetic-artistic meanings to forms with which world M appears to us, as usually is done with V. Why not to think at a "mechanical poetry" while observing a determinate machine language listing? Why not to attribute an æsthetic value to certain 0 - 1 sequences ? And what should we say about their precision, their density of sense? It's a matter of peculiarities of the machine (the individual, in this case): good expressions of its own machine-being. Moreover, in both cases (i.e. man or machine as individual), changing of interpretation (or computer, by analogy), starting from the same work (or program) Oi one arrives to a different world W in place of V (or N in place of M, considering the machine-individual). That is the expression of the multiplicity of meanings that become available with an Arp-like method of work, for example. More precisely, in this case such multiplicity is a consequence of the clean separation between syntax and semantics and of the possibility to redefine the semantic ties of symbols. Consequently, there're new interactions between symbols, meanings, operating methods. We are towards an abundance of sense possibility, but confusion is not a necessary result. In fact, the multiplicity of interpretative paths can constitute a higher artistic value, as it happens e.g. in P.Greenaway, where it assumes the form of a refined intellectual game.
The previously suggested parallel finds an authoritative philosophical support in Charles Morris' work, useful as epistemological foundation of programming languages theory too. Morris places semiotics as basis for the study not only of language but also of all human behaviors, Art included, of course. Its theory of semiosis, divided in syntax, semantics and pragmatics, is applicable also to the computer science universe in nearly literal sense; in particular all the aspects of efficiency, flexibility, ease of use of the computer system are subsumed in Morris' pragmatics. Thus he can be considered a forerunner of the modern theory of programming languages.
In conclusion, works (of Art) founded on the phrase of J.Arp exposed in article opening and those of a compiler can be placed, for many aspects, at the same level. Who ever of the two gets greater honor from this parallelism? In my opinion J.Arp: infact, while a machine cannot be interested in anything concerning J.Arp, the latter instead acquires - by analogy - the rigour that characterizes the work done by the automaton.