CONTENT
From philosophy of science (in german a bit more concrete ‚Wissenschaftstheorie‘, i.e. ‚theory of science‘) we can learn that a concrete science – like e.g. CS – has to be established within the framework of elements which constitute a scientific endeavour. Clearly, there are different opinions around what can be considered as a ‚framework of science‘. But there is a kernel of generally shared opinions which is strong enough to built on it.
(For a ‚condensed‘ information look to the diagram General Framework for a Scientific Program)
3.1) Starting point is the GROUP OF RESEARCHERS, e.g. the CS-Group. That group is able to communicate in a PRIMARY LANGUAGE L0 (e.g. ‚natural‘ english) which is embedded in a communicative context. This group has a ‚PRE-SCIENTIFIC‘ VIEW OF THEIR SUBJECT of investigation (in the case of CS is this view given through ‚candidates for semiotic processes‘).
3.2) The group is able to define in their primary language some METHODS OF MEASUREMENT. These measurements DEFINE THE DATA for a correlated FORMAL STRUCTURE, which represents the THEORY of the group. The data have to be represented in some formal representation, a DATA-REPRESENTATION-LANGUAGE L1. Furthermore has the INTENDEND FORMAL STRUCTURE to be represented in a THEORY-LANGUAGE L2 together with a formal meachnismen for INFERENCES. L1 could be a subset of L2.
Depending from which ‚point of view‘ one is looking to the ‚world‘ one can distinguish several main ‚types of measurement‘:
3.2.1) BEHAVIOR BASED (S-R := Stimulus Response Pairs)
3.2.2) INNER (PHYSIOLOGICAL; NEURAL) STATES OF A SYSTEM (N)
3.2.3) Correlation of BEHAVIOR and PHYSIOLOGY (S-N-R)
3.2.4) PHENOMENOLOGICAL (Ph)
3.2.5) Correlation of PHYSIOLOGY and PHENOMENOLOGY (N – Ph)
3.2.6) Correlation of BEHAVIOR and PHENOMENOLOGY (Ph)
3.2.7) Correlation of BEHAVIOR and PHYSIOLOGY and PHENOMENOLOGY (S-N-R-Ph)
3.3) The formal structures representing the theory are intended to ‚grasp‘ the ‚empirical content‘ of the subject by formal means. In any case this will be a form of APPROXIMATION. The whole unit of DATA and FORMAL STRUCTURE we call a THEORY.
3.4) It is possible, but not necessary, to set up an additionally formal structure with the aid of some additional formal language L3 to encode the INTENDED MEANING of the theory.
3.5) Besides the theory (and possibly a semantic) one needs an INFERENCE MECHANISM to generate possible consequences (THEOREMS) of the theory.
3.6) As in many cases today there will be the need for a COMPUTATIONAL MODEL of the theory. One reason is the complexity of the subject, especially if it has dynamic features. For a human brain it is nearly impossible to handle complex dynamic structures effectively. To claim those computational models with regard to the presupposed theory to be ‚scientific‘ one has to establish a MAPPING FUNCTION BETWEEN THEORY AND COMPUTATIONAL MODEL showing, which kind of relationship between both is supported. Insofar such a mapping function exists one can use the computational model to EMULATE the theory and thereby SIMULATE possible empirical substrates of the theory.
This is a general outline for ANY KIND OF SCIENTIFIC THEORY. We are interested in a special INSTANCE of this framework, namely what we are calling CS.
3.7) A scientific SEMIOTIC theory would be one which focusses itself onto processes which can be characterized as SEMIOTIC PROCESSES. If there are EMPIRICAL CANDIDATES and methods of MEASUREMENT a formal theory could be estabished, than we have a scientific EMPIRICAL SEMIOTIC theory. If we have no empirical interpretation but only FORMAL structures, we can speak of THEORETICAL SEMIOTICS or PURE SEMIOTICS. If one uses formal semiotic structures to set up any concrete thing/ object/ situation/ communication process/ clothes etc, one could speak of APPLIED SEMIOTICS. If one sets up algorithms/ computerprograms to emulate a {pure/ empirically interpreted} formal semiotic structure then one is doing COMPUTATIONAL SEMIOTICS.
Thus CS is defined by the existence of an algorithm, a given formal semiotic structure and a mapping function between both. To speak of computational SEMIOTICS instead of e.g. computational LEARNING THEORY, would crucially depend from the definitions of the presupposed theory: whether within the theory some definition – explicit or implicit – exists which is introducing the concept of SIGN.
Because in the history so far we have several differing definitions about what could be a SIGN (Peirce, de Saussure, Morris et al.) it is quite conceivable that there are several semiotic theories with DIFFERENT DEFINITIONS OF SIGN. It would then be one of the tasks of semiotic (META-SEMIOTIC) to analyze the similarities and differences of these concurring definitions and to make suggestions which variant should preferred under which respect before the other ones.
Here holds the same as in the case of the concept of a sign; lots of defintions are around dealing with the concept of KNOWLEDGE. Which one should be used and how this should be related to one of the possible definitions of a SIGN is a task on its own (EPISTEMOLOGY, THEORY OF SCIENCE, META-SCIENCE, META-SEMIOTICS etc.).
Today a huge amount of different types of FORMAL LOGIC SYSTEMs are kown. They are mainly used in the context of modelling and of constructing theories. Which one of these someone is using depends on his special goals.
From these types of logical systems one has to distinguis the IMPLICIT LOGIC of INTELLIGENT (SEMIOTIC) SYSTEMS. A biological system e.g. with a brain does all the time lots of NEURAL COMPUTATIONS which are realizing some KIND OF IMPLICIT LOGIC. The same yields e.g. for a consciousness which experiences the dynamic of phenomena showing some PHENOMENOLOGICAL REGULARITIES which are due to some IMPLICIT LOGIC. It has to be explored to which extend ANY SEMIOTIC SYSTEM is coupled to such an IMPLICIT LOGIC.
Another important aspect of intelligent as well as semiotic systems is LEARNING, the ability of a system to change its internal states depending from events within or without the system with some DURATION which is different from chance. Lots of learning concepts are around.
Much more KEY-TERMs are conceivable which can be of importance for a semiotic theory, especially also for CS.
