A termbase: Is not about the terms! (The concept-oriented approach)
Well, yes, a termbase is a repository of terms, but the central issue of termbases revolves around the organization of concepts. We have to differentiate between terminology work (the extraction of terms from sources) and termbase design (how terms are organized based on concepts). Once you identify the concept you look for its term. That is what they call in semantics and lexicology the semasiological approach: The grouping of words based on their meaning, or the concept-oriented approach.
Termbase creation starts by concept (or conceptual) analysis to clarify the definition of each concept and its scope or boundaries and identify its main characteristics to isolate it from other concepts. To identify concepts for analysis we need a concept system to organize them by subject field, and this is achieved by designing a concept map: a graphic representation of concepts and their relationships. Concept maps may be as simple or as complex as you want them to be, depending on the amount of information you are working with and the level of complexity of the termbase you are designing. In the concept maps you connect terms to each other based on the whether they are broad or narrow concepts. Here are one simple and one complex concept maps:
Concepts are structured in hierarchical (generic and partitive) or non-hierarchical (associative) relationships. Hierarchical relationships may be generic or partitive. In the generic hierarchy we find superordinate and subordinate concepts. The superordinate (also called generic or broader) concept is the general concept under which we have all the subordinate concepts that “inherit” characteristics from the superordinate concept (e.g. dog breeds: Labradors, bulldogs, Chihuahuas). Partitive hierarchy concepts are easy to identify because they are “part of” something (e.g. house, roof, door). In non-hierarchical or associative relationships, concepts do not share essential features, that is, they do not inherit characteristics (e.g. producer-product: bake-bread)
Concept analysis and mapping are very useful during termbase design because they help us come up with a clear structure for our termbase.
Sources and further reading:
- Pavel tutorial on hierarchies. Read here.
- Concept analysis. Read here.
- Wikipedia on semasiology.
- Semantic Relationships used in Controlled Vocabularies. Read here.
- Concept Systems for Terminological Analysis. Read here.
- The Theory Underlying Concept Maps and How to Construct and Use Them. Read here.