Christian Galinski
& Gerhard Budin
Infoterm, Vienna, Austria
Univerity of Vienna, Austria
Whenever and wherever specialized information and knowledge are created, communicated, recorded, processed, stored, transformed or re-used, terminology is involved in one way or another. Subject-field communication has become a specific type of discourse with specialized texts differentiating into a whole array of text types. When we define terminology as a structured set of concepts and their designations in a particular subject field, it can be considered the infrastructure of specialized knowledge. Technical writing and technical documentation are thus impossible without properly using terminological resources. Since the production of technical texts increasingly involves several languages, high-quality multilingual terminologies have become scarce and much desired commodities on the burgeoning markets of language and knowledge industries.
The research field we talk about is referred to as terminology science, its practical field of application is in terminology management, which includes the creation of subject-field specific terminologies and the terminographic recording of such information in the form of terminology databases, dictionaries, lexicons, specialized encyclopedias, etc. (For overviews and recent textbooks see, for English: [Fel84,PD85,Sag90]; for German: [FB89,AP89]; for Spanish: [Cab94]; for French: [Gou92].)
Concepts are considered the smallest units (atoms) of specialized knowledge. They never occur in isolation, but rather in complex conceptual networks that are multidimensional, due to a wide range of conceptual relationships among concepts. Given the limitations of natural language with regard to the representation of these concepts in specialized discourse (limited number of term elements in every language), concepts are increasingly represented by non-linguistic designations, like graphical symbols [GP95]. In addition we may distinguish between:
Theories of terminology as they have developed over at least six decades, consider concepts as:
The development of terminologies as a crucial part of special purpose languages reflects scientific, technical and economic progress in the subject fields concerned. Due to different speeds in this dynamic co-evolution of knowledge in the individual domains, specialized discourse continues to differentiate into more and more sectorized special languages and terminologies. But these communication tools become increasingly ambiguous, due to the sheer number of concepts to be designated and the limited linguistic resources of every natural language: terms are taken over from one domain (or language) into another, usually with varying meanings in the (productive) form of metaphors or analogies; new homonyms, polysemes and synonyms arise, motivating or even forcing subject specialists to standardize their terminology and harmonize them on the multilingual level in order to reduce and manage the constantly rising communicative complexity that faces their discourse communities.
But terminology research is not limited to comparative semiotic and linguistic studies of term formation and the epistemological dimension of the evolution of scientific knowledge. The agenda of terminology science also includes socio-terminological studies of the acceptance of neologisms proposed by terminology and language planners [Gau94], case studies on terminology development by standardization and harmonization efforts, research and development concerning the establishment and use of terminology databases for various user groups and purposes (e.g., translation, technical writing, information management) and concerning controlled vocabularies for documentation and information retrieval purposes (thesauri, classification systems, etc.).
Terminology management is primarily concerned with manipulating terminological resources for specific purposes, e.g., establishing repertories of terminological resources for publishing dictionaries, maintaining terminology databases, or ad hoc problem solving in finding multilingual equivalences in translation work or creating new terms in technical writing. (For terminology management see [WB95].)
Terminology databases are increasingly available by on-line query or on CD-ROM (e.g., TERMIUM, EURODICAUTOM), on diskette in the form of electronic dictionaries or as private databases established and maintained by engineers, computer specialists, chemists, etc. (working as terminologists, translators, technical writers) for various purposes:
For such purposes special computer programs have been developed (terminology database management programs), either commercially available on the international terminology market or developed as prototypes in academic research projects.
Due to the surprisingly high diversity of terminological resources that is potentially relevant to applications, terminology databases may look quite different from each other. One principle, however, seems to be the common denominator of all of them: the concepts under consideration are always the point of departure for database modeling; entries in terminology databases deal with one specific concept at a time. A terminological entry may contain not only term equivalents in other languages, synonyms, abbreviations, regional variants, definitions, contexts, even graphics or pictures, but also indications of relationships to other concepts (referencing to related entries) and subject-field indications by including thesaurus descriptors, class names from a classification system, etc., in order to easily retrieve terminological entries covering a certain topic.
The last few years have seen a considerable increase in epistemological studies in the framework of philosophy of science concerning the way in which scientific knowledge is constantly created, communicated and changed and the pivotal role scientific terminologies play in this respect. In the light of post-modernism, complexity, fractal and chaos theories, synergetics and other new paradigms that completely change our scientific view of the world and of ourselves, it is necessary to re-examine the correspondence between objects we perceive or conceive and the concepts we construct in the process of thinking (cognition, and re-cognition of objects) [DB88,Bud94].
The concept-oriented approach in terminology management mentioned above seems to be the key to solve a whole range of methodological problems in the management of multilingualism and information management in large international institutions as a number of innovate projects on the European level could prove [Gal94]. The performance of machine translation systems system could also be improved by integrating advanced terminology processing modules that are based on the conceptual approach to language engineering.
European research projects such as Translator's Workbench (ESPRIT Programme) or similar projects in
Canada (e.g., Translation Workstation), show a clear
tendency towards systems integration: terminology products are no
longer isolated and difficult to use, but fully integrated in complex
work environments. Automatic term extraction from text
corpora is one of the buzzwords in this type of practice-oriented
research [ADF
94]. A terminological analysis of
text corpora also includes fuzzy matching in order
to recognize larger segments of texts (complex multi-word terms, fixed collocations, but also semi-fixed sentence patterns).
Within the framework of the Text Encoding Initiative a working group (i.e., TEI A&I-7) has specifically been devoted to terminological resources and their management by SGML. Chapter 13 of the P 2 Guidelines of TEI on the application of SGML in text processing is dealing with the representation of terminological resources in SGML and the creation of an interchange format. This terminology interchange format (TIF) is now in the process of being standardized by ISO (ISO 12200, [MBW93]). The exchange of terminological resources has become one of the most discussed topics in the international terminology community. In addition to the introduction of the TIF standard, many methodological and legal problems (copyright, intellectual property rights, etc.) have to be solved.
Terminologists have also joined the international bandwagon of quality assurance and total quality management by starting research projects on how appropriate terminology management may improve the performance of quality managers, and vice versa, how to improve reliability of terminological resources by systematic quality management in terminology standardization in particular and terminology management in general.
The interdisciplinary nature of terminology science also becomes clear in its links to research in knowledge engineering and Artificial Intelligence Research [Ahm95,Sch93]. But terminological knowledge engineering (TKE) is more than just a series of projects and some new tools---it has also become a new method of modeling and representing knowledge in hypermedia knowledge bases serving as research tools for scientists (and completely changing research methods, e.g., by terminology visualization modules), as a knowledge popularization tool in museums, and as a teaching tool or as a hyperterminology database [IIT94].