The research group will tackle these questions from a number of different angles such as logic and philosophy, linguistics, computational theory, epistemology and the philosophy of science. This synergetic approach is new for communication theory. The results will be of interest for other disciplines, especially for the classical humanities and the cognitive sciences.

1.1 Motivation

Communication is important for many things; indeed, a certain amount is essential for survival. Without communication we wouldn’t understand the rest of the world – or even ourselves. Everybody knows this, and so do we. But this isn’t enough for us. We are inquisitive, we want to know more. For example as philosophers : When we know that communication is important, what exactly do we know? What is communication anyway? As philosophers of language : How are communication and language related? What means enable or at least simplify which types of communication in which languages? As epistemologists : How does our knowledge of what gets communicated come about? As philosophers of mind : Does communication presuppose language? And if so, what type? A public language – or only a language of thought? As computer scientists : How can forms of communication be modelled on computers? What can we learn about the design of computer systems from studies of communicative behaviour? And finally as human beings : Assuming we ever communicate again, what should we heed more than before?

Despite all the answers already given to these questions, much is still unclear to us. Moreover, we are convinced that new approaches will lead to better answers to many questions.

All of us have our own problems in understanding communicative understanding depending on our own special fields. Yet as we also share many problems (e.g. conceptual ones), we presume that many of our answers will be relevant for all of us. We believe that we can all answer our own questions more efficiently if we work together. And this is why we have set up this research group.

Top

1.2  Circumscribing the topic

Communication is a broad sphere – partly because the term "communication" is subject to such broad usage that it covers almost everything, i.e. everything between which something can be somehow transferred. Communication in this broad sense occurs for example between diverse fluid containers (‘communicating tubes’), between nerve cells, sensors and states of the brain, marital and other partners, television stations, computers, between humans and machines or other more or less complicated systems. The items exchanged during communication are equally diverse: Aids, broadcasts, containers, direct mail, energy to X and Y chromosomes, or zip codes. However, our interest doesn’t extend this far. Our research group is only interested in a certain area of communication, the area which is definitely most important for us humans (as humans) and most relevant for us as epistemologists, computer scientists, humans and philosophers, namely interpersonal communication . Furthermore, within this area our interest is limited to ‘transmissions’ which arise because we as the communicators want them to. The research project is primarily directed towards achieving a better understanding of communicative action . Thus by ‘communication’ we always refer below to communication by means of actions.

However, we are interested in the entire scope of communicative action. Therefore, we do not share the assumption of K.O. Apel, J. Habermas and others that ‘communicative action’ per se is a succession of actions by a number of people who all pursue the common aim of achieving understanding in the sense of mutual agreement concerning the existence (or the desirability thereof) of certain facts. Although we do not exclude this special case of communication from our considerations, it is not the main concern here.

Top

1.3  Different approaches to research – and our classification

As far as actions are concerned, we distinguish between the action execution and the action product (e.g. between the whistling and the whistle). This also holds for communicative actions and their general form: the communicative utterance of a linguistic expression. Accordingly, we initially differentiate between two research approaches, one of which is interested in questions of communicative execution, and the other in the specific characteristics of action products, especially linguistic utterances and expressions. Let us call examples of the first approach the Actors , and those of the second approach the Representatives . Both approaches are included in our research group, and neither of them ought to work independently of the other.

Our Actors include projects P1 to P3. Project P5 ( "Computational Dialectics" ) acts as a Representative in modelling dialogues, i.e. a specific Actor theme. Project P4 ( "Explanatory Coherence" ) belongs to neither the Actors nor the Representatives. To be fair, the members of this group should be given a name as well, so let’s call them the Coherentists . Their work is badly needed by the Actors, the Representatives and the envisaged (see this §, (C) below) Mentalists, for all three experience coherence problems which call for a systematic solution based at a deeper level. And pinpointing this solution will be next to impossible without the help of the Coherentists.

Top

As is always the case with action theories, communication theories can too be subdivided into three types of theories. Empirical communication theories study how communication – in individuals, and within certain groups, strata or cultures – actually takes place. By contrast, normative and rational communication theories address how communication ought to take place. Furthermore, we must always distinguish between the semantic (conceptual) and nomological component of theories. The nomological component of communication theories concerns the formulation of laws and principles of communication with respect to content (empirical, normative or rational), whereas the conceptual component deals with explaining a language which is sufficiently expressive and as precise as possible, with which the sentences of the (empirical, normative, rational) communication theories can be formulated. Although this language need not be the same for all communication theories down to the last detail, it should be in essence. Formulating and substantiating this essence is the job of a General Theory of Communication .

The goal of our Actors is not to develop an empirical or a normative theory. And with the exception of questions of linguistic correctness, the same goes for our Representatives. By contrast, rationality assumptions play a major role in both; this is especially apparent in the projects of our own Actors. Just how this role can and may usefully extend is, however, itself one of our most contentious issues. It must never be allowed to escape the attention of the monitor, project P3 ( "Speech Act and Interpretation" ).

The Actors will analyse key aspects of problems arising during the construction of a General Theory of Communication , which will be immediately dealt with by the philosophy projects, P1–3.

Top

Compared to the Actors, the Representatives have the advantage that they go back immeasurably further. The entire Occidental semantics of Plato to the present was and remains almost exclusively representative. According to this tradition, linguistic expressions are significant in that they represent something: things, quantities of things, quantities of quantities etc. Most of our formal logics are based on this (also known as realistic, correlation-theoretical or representation-theoretical) semantics, the framework of which has by now also been precisely described in the form of various models. The Representatives are thus situated at a higher level of theory development than the Actors.

Consequently, a map able to pinpoint our Representatives would have to be much more detailed than the above description of the current Actors’ setting. It appears in the individual application for P5. At this point we shall limit ourselves to some of the most important aspects of recent developments relevant to a general theory of communication.

The relevance of non-classical logics has also been acknowledged for diverse application problems in computer science, especially artificial intelligence. In particular, non-monotone methods of inference based on rules with exceptions have been studied, and such methods of conclusion have nowadays been formally very well documented. Nevertheless, as considerable problems of complexity arise, we are still far removed from the original aim, namely modelling ‘common-sense reasoning’ on computers. Project P5 ( "Computational Dialectics" ) will attempt to move ahead in this field by modelling procedural aspects of truth ascertainment and decision-making in a rule-based dialogue.

Top

(C) The main focus of our propounded Mentalists

One specific mentalist project is not yet part of our group; nevertheless, we are striving to extend our work towards intentionality via the addition of such a group. As Actors , our Mentalists will then for instance study some of the preconditions which must exist for successful communication to take place between two ‘systems’ (people, different living things, possibly also machines). The aim is to establish the extent to which genuine , intrinsic wishes, needs, assumptions of belief etc. must be present on the side of the ‘speaker’ and the ‘listener’. By contrast, as Representatives they will among other things have to deal with the question of the manner in which what gets communicated gains meaning or intentionality extending beyond its syntactic content.

The Actors and the Representatives encounter coherence problems at every step. After all, like any other action, a communicative action is only understandable if it ‘fits’ its environment. And exactly what belongs to the content of a linguistic utterance depends on how this content would ‘fit’ utterances having the same linguistic form. New information often appears not to ‘fit’ old information, without one directly contradicting the other. Etc. The problem is simply that no-one really knows yet what ‘fit’ means here and/or there. The Coherentists will be expected to give practical support to all sides running up against this problem (i.e. with a ‘fitting’ coherence theory) – which does not, however, imply the expectation that the Coherentists’ help will also be able to eliminate all specific problems of ‘fit’.

Top

2  The state of research

The state of research is outlined in the sketches of the individual projects. This survey contains just a few key elements important for the classification and the context of the projects.

In order to be able to speak about actions more accurately, our distinction between the execution and the product must be augmented by further differentiation: between types and occurrences . This distinction can be applied to the actions themselves (the executions) and to their products. This results in the following differentiation diagram:

	Act	Product
Type	Mode of Action	Form of Action Product
Occurrence	Concrete Act	Product of Concrete Act

and accordingly for linguistic actions:

	Act of utterance	Product of utterance
Type	Form of utterance	Expression
Occurrence	Utterance (speech act)	Concrete Expression

As already mentioned, the main focus of our Representatives is on the right-hand side of the diagram, the product side, especially the linguistic expressions. But although the main focus of our Actors lies on the left, the question is, where exactly?

Top

a) The state of research

The Actor is faced by the following fundamental decision: How should, and indeed, how may one set about constructing a General Theory of Communication in the first place? By using concrete action concepts (meaning concepts which are initially only defined for concrete acts) or act-type concepts? And in particular, what action sense is primary? The subjective sense (the sense which an action subject personally associates with an act) or the intersubjective (the sense which an act has within the group concerned)? These questions harbour the following fundamental problem: In explicatory terms, what comes first – the individual or the group? (This is a shorter form of the more precise question: What concepts are primary: those which are initially only defined for individuals, or those which are initially only defined for groups?) In the social sciences, opinions over this question are divided between the individualists (subjectivists) and the intersubjectivists (collectivists).

This dispute is relatively new in the theory of communication and in a theory of language orientated towards communication. Outlines worthy of serious discussion have only been possessed by both sides since the mid-20th century: by the intersubjectivistic side since Wittgenstein’s practical theory of meaning, and on the individualistic side since H.P. Grice’s first sketches.

Although both drafts were published at about the same time (1957/58), their reception was very different. Until well into the 1980s, discussion was exclusively dominated by Wittgenstein’s advocates; most of the philosophers supporting Grice are still personally known to one another. By contrast, linguists who profit from Grice’s ideas are nowadays two a penny.

What has become of these two drafts? The action aspect of Wittgenstein’s idea of identifying the meaning of signs with their usage in the various language games was conceived within the framework of the speech act theory conceived by J.L. Austin and then mainly developed by by J.R. Searle (or to be more accurate, as a special part of the theory of illocutionary acts ). Wittgenstein’s goal of fathoming the entire linguistic meaning by determining usage was thus, despite claims to the contrary, de facto abandoned. W. Alston was the only one to sketch (on less than five pages at that) what a practical theory of meaning based on language theory could look like. It was only in 1983 that E. von Savigny developed this sketch into an extensive theory. However, his writings also contain the sharpest objections to Grice’s programme. Related ideas can more recently also be found in R. Brandom’s work.

Grice’s programme comprises three steps:

1. First explain a general concept of communication using action-theory terms (individualistic of course!), i.e. a concept of communication which does not yet make use of concepts of conventional or linguistic meaning ( action-theoretical foundation of communication theory ).
2. Then show how the concepts of conventional or linguistic meanings can be determined with the help of the general concept of communication introduced and thus in the end can again be determined using action theory ( action- or communication-theoretical semantics ).
3. And finally explain the diverse cases of so-called indirect communication, in which what the speaker gives to understand differs from what she actually says ( theory of implicatures ).

Step 2 can simply be skipped by accepting another type of semantics and just working with 1 and 3. This has actually been done by most of those using Grice, including nearly all the linguist users.

The state of play of the Grice programme is as follows: Step 1 is largely complete. A general, systematic outline is not yet to be observed among the published books or papers on 2. Although the literature concerning 3 is enormous, the question of what exactly should be understood by implicatures has so far remained almost completely omitted. There are no comparisons whatsoever between the individualistic and the intersubjective approaches in terms of their strengths and limitations.

b) Special Actor interests and intended progress

Projects P1 and P2 will clearly follow without any ifs or buts the individualistic viewpoint. P1 ( "Reconstructing Speech Act Theory" ) considers expansion. How much of speech act theory can also be reformulated within the framework of an individualistic communication theory? Knowing this is important for assessing the individualistic programme as a whole, as this capacity for reformulation – in connection with Step II of the Grice programme – is something like a test case for the strength of the methodological individualism propagated by Max Weber.

Project P2 (" Pragmatic Implications" ) will tackle for the first time Step III of the Grice programme in a strictly systematic way. Our main goal is to develop a theory of implicatures by means of treating implicatures as a subtype of communicative acts. Only once this project has been completed will we be able to adequately assess the large quantity of applied philosophical and linguistic works on so-called implications. One main problem of this project will be distinguishing implicatures from other pragmatic implications (especially non-semantic presuppositions).

Project P3 (" Speech Act and Interpretation" ) will examine the individualistic paradigm – and confront it with the programme and the initial elaboration of an opposite reconstruction which elaborates approaches of formal dialectics (N. Rescher, R. Brandom) for the philosophical analysis of communication. This will be done in the framework of computer science by project P5 (" Computational Dialectics" ).

(B) Concerning our Representatives

In order to model everyday inferences, use is now also made of non-classic logics in artificial intelligence . Numerous problems in the fields of action modelling, planning, temporal inferring or for example model-based diagnosis entail representing typical circumstances which normally apply – but not exclusively. Various non-monotone logics have been developed to represent rules with exceptions. Although these logics lead to an acceptable, more precise specification of a non-monotonous concept of inference, they are unable to explain the performance of our everyday inferring. Quite the opposite, in fact, for complexity analyses show that inference techniques in these logics are possessed of extreme complexity.

Techniques of regulated communication are frequently used in practice which are supposed to enable a specific question to be answered or a decision to be made within a certain predefined space of time. Completeness and correctness are overshadowed by fairness and efficiency. One classic example is court hearings: although each side gets the chance to present its point of view, in the interests of efficiency it cannot go on presenting new arguments ad infinitum . In particular, care is taken that irrelevant arguments presented merely for tactical reasons are not allowed to hold up the proceedings.

Project P5 will try to represent such techniques of regulated communication in order to arrive at realistic and efficient computer models of argument and decision processes.

(C) Concerning our propounded Mentalists

There are numerous links between the philosophy of mind on the one side and communication theory and philosophy of language on the other. This is not surprising, since our language is regarded as the mind’s most important instrument and means of expression. Accordingly, the various paths of development of communication theory and the philosophy of language also reflect the prevailing viewpoint of the philosophy of the mind. This is especially apparent in the 20th century. Behaviourism , which overcame the older mentalism, identified mind with behaviour . The subsequent structuralism identified the mind with innate or upbringing-related individual or collective depth structures – and language quite consistently as a special type of behaviour or as a special form of such depth structures with its own characteristic rules . Wittgenstein’s later philosophy can thus also be regarded as a synthesis of these two approaches, regarding our behaviour itself from the viewpoint of rules; and mind can only be ascribed to those who follow the rules anchored in the various language games. According to this approach, mind presupposes language, and not the other way round. Yet precisely this is disputed by mentalism , which has re-appeared in both behaviour and also communication and language theories. The mentalistic approach currently dominates the debates in the philosophy of the mind. (Working out any restrictions connected with this is one of the topics of project P3, "Speech Act and Interpretation" .)

Despite their fundamental importance for inference techniques and epistemic justifications, there are not yet any really convincing explication proposals concerning what characteristics make a belief system especially coherent. The most promising are the links with the explanation concept, which for its part adopts a key role in many debates on the philosophy of science. Explanation relations can weld our opinions together to create a system. Exactly how they do this and when a good explanation exists will be established by project P4. We will also have to enter the territory of debate in the philosophy of science concerning the explanation concept – for this is where the preliminary decision regarding the suitable explanation concept will be taken.