Word Formation in Electronic Dictionaries

Word Formation in Electronic Dictionaries Pius ten Hacken In this article I will discuss various aspects of the relationship of word formation to dictionaries. Although the focus here is on electronic dictionaries, the discussion will also touch on certain types of dictionaries for human use, especially learner's dictionaries. I propose that an approach diverging significantly from the traditional treatment of word formation in dictionaries has decisive advantages for the development and use of electronic dictionaries, and that some of these advantages carry over to the treatment of word formation in learner's dictionaries . In section 1 the problem is explained. Section 2 is devoted to a review of some approaches that are modeled on the traditional division of labor between dictionaries and grammars. In section 3 the approach incorporated in Word Manager is presented, which has a number of advantages compared to the approaches discussed in section 2. Section 4 outlines a way of using these advantages in learner's dictionaries . Section 5 summarizes the conclusions. 1. Introduction One of the most common ways for the vocabulary of a language to be extended is by the creation of new words by word-formation rules. Word formation is traditionally divided into derivation, e.g., forming regularize from regular, and compounding, e.g., forming bookshop from book and shop. Word formation as a whole, considered as a component of interacting rules, is very productive. A rule such as the one adding -ness to adjectives can apply to almost any adjective, and the adjective itself may be the result of another word-formation rule, e.g., adding -ful or -less to a noun (fruitful ) . The meaning of such words is largely predictable on the basis of their components, but various de- Word Formation ¡? Electronic Dictionaries1 59 grees of irregularity may arise once the new word exists. Thus, textbook and modifier have specialized senses that are more specific than can be predicted on the basis of the rules and the words to which they apply. 1.1. Electronic and other dictionaries In dictionaries for human users, word formation is usually not seen as a major issue. There is an almost general consensus that can be summarized as follows: it is impossible to achieve completeness because of the productivity of word formation and at the same time unnecessary to aim for it because of the regularity of the new words. Of course irregular cases should be treated, but there is no need to treat a compound like textbook as being any different from simple words such as textile. As far as the relationship of text and book is relevant, it is reasonable to expect that any user will immediately recognize this relationship . Regular cases such as yellowness are typically given as run-on entries without definition, considering space limitations. This treatment is legitimized by the assumption that human users of the dictionary know the word-formation rules of their language and can apply them even without being consciously aware of them. Electronic dictionaries are different from dictionaries for human users primarily because they are intended for use by computers in systems developed in computational linguistics (CL). Examples of CL systems are grammar checkers, machine translation systems, and dialog systems giving access to information encoded in a database. In general, CL systems are programs that make the computer perform certain tasks dependent on human language. The properties of computers differ greatly from those of human dictionary users and in order to fulfill the users' needs, electronic dictionaries should be set up quite differently from dictionaries for human users. The basic differences are that electronic dictionaries are more explicit and more specialized. Explicitness is required because only information expressed in a formalism the computer can read is accessible to the CL system. Specialization is required not only for areas we find in human dictionaries, e.g., language and subject area, but also for the task of the CL system, the theory used in the approach to this task, and the formalism with its exact interpretation. As a consequence of these differences, two important issues in the production of electronic dictionaries are reusability and consistency . Specialization results in an increasing number of different die- 160Pius ten Hacken...