8'th occurrence of 'formantics' on the Internet

by Karel Thönissen

Newly minted: formantics

In this article, I used the word formantics for the first time publicly. I minted the word about 6 years ago, but until now never used it in public writing. Google only has 7 other entries for the word, and these are entries for meanings that are distinct from mine. So the word exactly has the meaning that I give to the word. Humpty Dumpty, anyone?

But what does it mean?

Stamper's semiotic framework distinguishes 6 levels: physical, empiric, syntactic, semantic, pragmatic, and social. Signs can be studied from each of these perspectives. I refer to Stamper's writings for more details, and concentrate on the difference between the syntactic and the semantic. Syntax studies the valid appearances of signs, e.g. with grammars. Semantics studies the meaning of signs, i.e. the relationship of signs and things they stand for. Usually these things they stand for are not signs themselves (although theoretically, anything can acts as a sign for something else), although that is certainly possible. However, if signs refer to other signs, they at least do so at a more concrete layer.

However, when computer scientists talk about semantics, they mean the effects of the execution of a sign (the computer program) in terms of another sign (the computing environment). From a semiotic point of view, these signs are equivalent, and neither of the two is semiotically closer to the meaning of the sign. What computer science calls 'semantics', are in fact rewriting rules or transformation rules for signs. These rules do not relate signs to their meanings. Therefore, semiotically, this is a branch of syntactics, often called formal semantics.

This use of 'semantics' ('a branch of syntactics that is called formal semantics') is so confusing if you are part of both communities, that I minted my own word. In my semiotic writings, the semiotic framework has 7 layers (that have nothing to do with the evenly numbered OSI-model): physical, empiric, syntactic, formantic, semantic, pragmatic, and social. Formantics is the study of formantic properties of signs.

Finally, as an illustration of the difference, take this part of a Prolog-clause:

• eat(A., time1), eat(Karel, time1), doDishes(Karel, time2), sing(Karel, time2), after(time2, time1)

Asked about the formantics, I could explain how expressions like these are evaluated, i.e. converted in other expressions, possibly involving the sign that is the concatenation of all bits in a certain computer and the devices attached to it. I would talk about algorithms, equivalences, equalities, and rewriting rules. Under certain formantics, the expression above would be equivalent to:

• eat(Karel, time1), eat(A., time1), doDishes(Karel, time2), after(time2, time1), sing(time2, Karel)

However, if you asked me to explain the semantics of this expression, you would have to invite me at your place for a meal. Then I would eat with you, and do the dishes afterwards whilst singing. Feel free to invite me, I might consent....