It is now recognised that the definition is not alone in providing a link be
tween term and concept but is supported by the declaration of a sufficient and appropriate set of conceptual relationships. The definition can therefore concentrate on the essential characteristics a concept has in common with others and which differentiate it from other concepts; by contrast the relation
ships indicate the type of link a concept has to other concepts in the system.
These two complementary types of information assist in the identification of equivalent terms of the same or other languages and can also be used in
case the meaning of a concept is to be restricted with a view to its use in a
standardised form. ' This dual identification of concepts may in future be formalised so that the
natural language formulation of the conventional definition can be replaced by a listing of characteristics only.
THE INTERACTION OF RELATIONSHIPS AND DEFINITIONS
Conceptual Definitions Terminological Variants system (C) (D) system (T) (t) [C1] [Dl] [Tl] [tl]-[tl]
\ I I / I
[C2] [D2] [T2] /
i \ / I
[C3] [D3] [T3] [t3]-[t3]-[t3]
[C4] [D4] [T4]
[C5] [D5] [T5] [t5]
The definition links concepts to terms (expressed horizontally in the diagram), but concepts and terms are themselves linked to each other (expressed by vertical and diagonal lines in the diagram). Terms are further linked to their variants and variants themselves may be linked through their form.
Terms can be classified and presented in a thesaural structure. The flexibil
ity of data selection and presentation in term banks permits displays of terms surrounded by their broader, narrower and related terms. The usefulness of this type of display has been demonstrated in documentation thesauri but also in such pioneering lexicographical works as Wüster's Machine Tool Diction
ary (1968). There is also a tendency to construct documentation thesauri so as to permit consultation for terminological purposes, e.g. the new ROOT Thesaurus of BSI (1980) and the Construction Industry Thesaurus (1976).
The relationships among terms contain elements of definition as shown by such simple statements as the following which underlie the symbolic repres
entation of Broad Term, Narrow Term and Related Term.
x is a type of Y x is a part of Y x is related to Y
The Cognitive Dimension 53 Such statements can be generated automatically by the computer which con
verts this kind of information into a simple definition. It is equally possible to generate automatically expanded defining statements such as:
x is a type of y together with z and q x is a part of y; other parts are z and q x is related to y and so are z and q
Moving further up, down or diagonally across the relationships expanded statements, giving, e.g. the top or bottom terms of a hierarchy, are equally possible.
Partitive and generic relationships are not the only types of relationships and others, such as the causal or genetic, can be added, thus diversifying and adding to the basic information required in definitions. Property terms can be associated with the entity terms they qualify, process terms can be associated with the products which are involved. In this way differentiating characteristics between generically related terms can be associated with the narrower specific term.
If systematically pursued such more detailed terminological identification corresponds to the requirements of the terminological definition described in the previous section.
A close link can thus be established between definition and placing a term in its environment, and defining a term can be considered to be synonymous with fixing its identity in relation to other terms.
The practical adoption of such new types of information categories in term banks would, however, require clarification of a number of issues currently insufficiently researched. They are here expressed as a number of questions to which answers would have to be given before progress can be made with new information types.
1. The nature and types of terminological relationships.
- Beside the well-known relationships are there others common to all sub
ject fields?
- Are there relationships particular to certain subject fields?
- How can the relationship between horizontally related term be specified?
2. The identification of terminological relationships.
- Can simple techniques be devised for accurately identifying relation
ships?
- How many different relationships can a terminologist be expected to identify accurately?
- How can one achieve consistency of attribution among different termino- logists?
- What computational techniques can assist in this task?
- How can the computer be employed to control consistency of work?
- What is the minimal information required for the computer to generate a full representation?
3. The presentation of relationships to users.
- Which defining patterns can provide the maximum amount of relevant information to all specialised users?
2.4.1 Complementarity of definition and relationships
The process of defining as performed by terminologists is conveniently seen as an extension of the work involved in establishing relationships among terms.
The equation of these two operations has a number of theoretical and practical advantages. On the practical side it maintains the unity and consistency of the identification of a term within the knowledge structure where it is used.
On the theoretical side it represents an acknowledgement of the fact that many terms, in particular those for which no definitive external definition exists, undergo small variations of meaning which impinge on the intension of other terms and would thus invalidate a highly restrictive definition.
It is the recognition of varying degrees of elasticity of the knowledge spaces occupied by some types of terms which suggests that the static approach of definition is unsuitable as it tends to ignore the dynamics of the segmentation of the whole knowledge structure.
It is therefore the type of term which determines the need and the type of definition required. Definitions should never be allowed to restrict the flexibility of natural language and seeing terms as variously strictly defined allows for the innovative use of language in harmony and not in conflict with the control mechanisms inherent in term banks and dictionaries.
Chapter Three