In the current debate about the status of lexicography there are at least three quite different opinions:
1. Lexicography does not have and does not need any kind of own theory but can use all relevant linguistic theories;
2. Lexicography needs a theory special for the lexicographical praxis, but this discipline is still a part of linguistics;
3. Lexicography is a genuine part of information science and can use theories and learn from practice in the information society, but it also needs special theories for lexicography.
It is the last standpoint we will follow in this paper discussing the information needs in the information society partly using the function theory of lexicography. The paper will briefly address issues regarding information overload and information stress showing how commercial systems try to address this by means of relevance ranking based on system relevance. It is argued that this is insufficient as system relevance does not take into account the unique situation of an individual user in a specific information need environment. We argue that a user has different information needs over time in a specific contextualised environment such as a work task. A user profile can be constructed on the basis of a number of variables such as the user’s knowledge base, language expertise in terms of the subject discipline and the language of the document (L1, L2 etc.). Even though certain aspects of a user’s profile remain (fairly) constant (e.g. L1 proficiency, domain knowledge) his/her profile can change in terms of the information needs depending on the specific situation. For example, a user may usually require detailed information aimed at the expert if he/she needs information for a standard work task in which he/she is an expert; however, if the user needs to read background information in a discipline in which he/she is only a lay person, the nature of the information needed to satisfy the information need may change to less detailed information aimed at a lay person. The available information can therefore also be characterised in terms of a number of variables, such as complexity, detail, language etc. To satisfy a user’s information needs there should be a match between the user’s profile in a given situation and the available information. This will be discussed in more detail and illustrated by means of examples.
Dictionaries as information tools tend to provide as much information as the lexicographer deems necessary, without taking into account the actual information need of a user in a specific situation. Within the function theory of lexicography it is suggested that the information provided be filtered in terms of the situation for which the user needs the information, viz. a communicative situation (either text reception or text production) or a cognitive, interpretative or operative situation. The differences between these situations will be discussed briefly. We will show at the hand of examples how this approach to lexicography is transforming dictionaries as information tools in the dictionaries of the Centre for Lexicography (Centlex) at the University of Aarhus.
The principles and theory underlying the Centlex dictionaries therefore make provision for filtering data in terms of the specific user needs in a given situation. The remainder of the paper will show how modern information technologies and techniques can enable the user to customise the information made available at a much more granular level based on the specific information need in a given situation. Technologies and techniques include advanced search and filtering options, adaptive hypermedia, detailed markup of data in the database by means of metadata in an RDF schema and open and linked data. The user is empowered to customize the granularity of the information that he/she accesses in any given information need situation and remains in control of exactly how much or how little information is made available to him/her by the set of e-information tools.
We propose that predefined filters be provided for the most common user situations (as is currently the case in the Centlex dictionaries) and that users with different information needs be empowered to use another search order, to do searches in more/different fields, to have the data presented in a different order or to have data presented from different/additional fields, depending on their specific information need in a given situation. In the end there would be the possibility of an unlimited number of different dictionaries based on the customization principles discussed above and the number of fields in the database. A good tool is a tool which is able to fulfil the needs of a certain user group by giving quick access to the data, by giving relevant and correct data in a way which is understandable for this user group, but which does not send the user into information death by giving a large amount of unneeded data. Such a good tool must necessarily be a monofunctional tool prepared for a certain user group with a certain type of information need according to a certain type of user situation.