Interactions (SPILE)

Summary

In this project attention will be paid to social and philosophical issues interacting with other projects within the coordinating Parlevink project. These issues can be subdivided into social aspects of linguistic engineering, philosophical aspects of artificial intelligence and methodological aspects and pragmaticism.

Objectives

Within this research project we will pay attention to the interaction between social sciences, philosophy and linguistic engineering. For the time being this theme has been subdivided into three more concrete and specific topics:

• Societal Aspects of Linguistic Engineering;
• Philosophical Aspects connected with Artificial Intelligence;
• Methodological Aspects and Pragmaticism.

In contradistinction to the first topic which is social science and society oriented, the other two topics have as a common focus the search for the distinguishing mark of linguistic engineering within the area of information technologies.

The first topic concerns the interaction between linguistic engineering (sometimes implicit as a part of information technology) and its social environment. Concentrated on questions concerning the responsibility of the language engineer (computer scientist respectively) in this context, concrete subjects of investigation in this area are:

• the responsibility of computer scientists as viewed in retrospect from the historical development of calculating machines, computer technology, and artificial intelligence;
• the choices that can be made within the development of the Schengen Information System, as an example of an information system with a huge impact on the private life of citizens;
• drawing up an inventory of the variety in existing research policies with respect to linguistic engineering; evaluation of e.g. purposes, aims, expectations, promises and presuppositions that play an important role in the steering of this research area.

From this research topic, assistance has been given to the organisers of the LEC'93 symposium on "Linguistic Engineering in Context", held at the University of Twente.

The second topic is concentrated on the question what makes the great difference between common language use and computation such, that the latter process can be executed by machinery far more easily than the former one. Is it that within linguistic engineering we meet the borderline of real intelligence? When we put the question in this form, we have to declare what (real) intelligence is. One can ask which specific human functions are called intelligent and why people feel threatened when confronted with a technological realisation of these so-called intelligent functions. In this connection the prevailing tradition of materialistic reductionism and the presupposed mind body dichotomy have to be questioned. The main issue will be to explain why the cut of Cartesian dualism offers an adequate theory for language engineers, and how a pragmatistic framework may help to overcome the less rigid dichotomy between body and mind, relating intelligence to both of them.

The third topic takes it point of depart in the observation that within linguistic engineering we have met the limits of the paradigm, up to now quite dominant within information technology; the paradigm of the unrestricted application of general and universal formalisms to the objective representation of information. On methodological grounds it will be shown that this paradigm has to be replaced by a pragmaticistic semiotic one. The latter is founded in the relationship between formalisms, actual representations of knowledge, and the functional or contextual aspects linking formalisms and representations. Within this new paradigm, it is essential that representations are based on purposes, contextual habits and meaningful contexts. It turns out that the need for this paradigm has not only to be found within linguistic engineering, but also in the more traditional branches of information technology. The latter will be shown as a spin-off of the Schengen project, mentioned before. In this connection the specific relationship between the representation of information and the usage of this information in public administration will be investigated.

Motivation

Within the philosophy of technology it is widely accepted that the stage of information technology is distinguished from previous stages of technology, because of a different relation between the functionality of the tool and the control of the user. The impact of this philosophical statement has not led to a more general theory of design and development of computer systems so far. The usual approach to computer systems does not differ very much from the approach to instruments or to machines, that although, of course, the user must be enabled to manage the tool in question, the functions to be afforded by the tool may be specified independently from it. In both the philosophical as well as the technological conception of tools, this approach leads to a dichotomy in the understanding of the relation between tool and user. The first view, the instrumental one, is that the user is totally free in the decision of using the tool and, if so, for what purpose; the second view, as can be found with some classical philosophers of technology as for instance J. Ellul, is that machinery has totally determined the social setting in which decisions concerning the use of tools and purposes have to be made. Neither of both approaches will fit the relationship between computer systems and their usage completely.

Especially within the field of linguistic engineering the classical borderline between the type of functions to be implemented in the computer system itself and the type of functions realised in the actual use of the system becomes less distinguishable. In developing natural language models and systems one has indeed to implement some sophisticated features which are presupposed in any non-trivial human-computer dialogue. For instance, the "relevance" of the different phrases in a dialogue has to be taken into account and everything will have to be analysed within a dialogue-specific, goal-directed framework.

The intention of the Parlevink project is that the relationship between the development, the design, and the application of computer systems on one hand and their historical, ethical and philosophical aspects as well as their impact on society on the other, will be given attention to. Obviously, the point of view of the investigations is mainly defined by language processing. The history of information technology, the role of a (linguistic) engineer in society, the characteristic properties of computer science and information technology are among the issues that will be studied.