Meta-Dictionary

A language is not a uniform or one-dimensional system, but is marked by constant change and variation, not only in terms of its historical development but also, and especially, in its current form. Alongside the "standard" language, there are regional variants, group languages, specialist languages and much more. Even a native speaker can easily lose track of all the different permutations of his or her language, and for a foreign learner, the task is even more daunting. Dictionaries and reference works of all types are therefore essential aids, for they list and order the diversity of the language, and also provide the most comprehensive and efficient access to language information.

There are many different types of dictionary which cater for the different variants of language and users' wide-ranging needs and contain carefully researched information. They thus remain key tools for research and study, for the data collected by the Internet search engines cannot achieve the reliability, systematisation or level of detail provided by reference works, and therefore still lag far behind in terms of precision and scope.

The University of Trier's Centre for Electronic Retrieval and Publication Methods in the Humanities digitilises printed dictionaries for publication on the Internet or on CD-ROM. This includes standard reference works such as the German Dictionary by Jacob and Wilhelm Grimm, extant Middle High German dictionaries, several dialect dictionaries, and the Goethe Dictionary. These dictionaries can be used as self-standing works and researched according to various categories, but they are now being integrated into a uniform digital system – the Dictionary Network.

Each work is encoded according to its specific requirements, which means that as well as indexing straightforward textual information, added value is created through the provision of additional content in the form of meta-data: details of a dictionary entry such as keywords, grammatical information, meaning, supporting quotations, etymology, references etc. are marked up so that they can be

In order to create integrated access to the various dictionaries, which have different structures and use different keyword approaches, a single standardised keyword, known as a 'hyper-lemma', can be used, enabling dialectal, historical or standard-language keywords to be located.

By categorising dictionary sources by means of symptomatic values, additional opportunities are created to search the material according to region, time period, text type, style, etc. Together with the classification structure described above, this means that in the Middle High German dictionary network, for example, complex searches can be undertaken – for example, to locate all derivations with the suffix "-unge" in religious texts from Bavaria in the 13th century. Classification of the content of reference works also adds value by creating new indexing and search options through access to a hierarchal, thematic classification system, such as the Dewey Decimal system, which is a standard system in use across the world. For example, in J.G. Krünitz's Ökonomische Enzyklopädie, it is possible to target and locate all articles dealing with botanical museums and collections by following the link 'natural sciences > plants > museums'; tracking by mouse brings the user to the desired article.

Digitalisation creates the key prerequisites for further networking: dictionaries reflect the diversity and heterogeneity of the language but are also linked in many different ways, as is apparent from their explicit references to each other. These references are initially shown as monodirectional hyperlinks, but digitalisation enables multidirectional references to be created as well: this means that older dictionaries can be made to refer to more recent works. For example, online, it is possible to create a reference in the Rhenish Dictionary of 1928 to the more recent Dictionary of the Language of the Palatinate, published in 1965. Such a reference could not created in the printed work.

Even without explicit references, dictionaries are linked at many levels, as they contain information about the same or similar themes and cover the same topics, albeit from different perspectives. Nonetheless, an overarching integrated search is not always possible due to the differences in the design, order and structure of the individual works. Digital reference works that have been structured and indexed as described above can make these substantive links explicit and visible. For example, via a hyper-lemma, the user can search for all variants of the German word 'Brombeere' (meaning blackberry) (e.g. Brombeere in modern German; brâmber (Middle High German); Brame, Brambeere, Frambeer (Rhenish); Bromelen (Letzeburgesch) etc.) or for all possible words meaning "blackberry" (e.g. Brombeere (modern German), Schmärzbeere (Rhenish), Schmääspel (Letzeburgesch), More (Moselle Franconian), etc.).

This enables a new quality of information-gathering to be achieved, closing the gap between the inconvenience and limited availability of the printed versions, on the one hand, and the unsystematised and random nature of the information on the Internet, on the other.

The ambitious goal, outlined here – namely to create an intelligent "meta-dictionary" of the German language with links on many different levels – is still in its infancy and can only be achieved through joint research and practice across a range of disciplines. Concepts, methods and procedures from the various areas – from IT to computer philology/linguistics, systemic linguistics and lexicography/lexicology – must join forces to create a set of interdisciplinary tools for this purpose.

The Dictionary Network not only offers users innovative and efficient access to linguistic information. For students of the German language, it is an exciting and inspiring portal which may tempt them to sample, browse and delve into the wealth of dictionaries that their chosen language has to offer.