|
Current working practice of established German dictionaries incorporates large corpora as the basis of most analyses, descriptions and presentations. It is, however, individual lexicological and/or different corpus-methodological approaches that play a crucial role in the process of extracting and documenting lexicographic information in individual reference works. This paper addresses the question of how reliable information is in some electronic German dictionaries. Objects of our investigation are different types of corpus dictionaries, e.g. a digitized dictionary, a reference work that compiles its data fully automatically, a lexicographic system combining different electronic resources, and a corpus-assisted dictionary that examines and interprets its corpus data lexicographically. Critical examinations of such reference works inevitably come up with questions of authenticity and reliability of the given dictionary information. The advantages and disadvantages of various lexicographic or corpus-linguistic methods which are individually implemented will be outlined and critically analyzed with the help of examples. According to an extensive study (cf. Müller-Spitzer 2011) reliability of given information is one of the key criteria assigned to any reference work by users. We will elicit how different corpus methods expose different descriptions of natural discourse and how they answer questions of authenticity, typicality and reliability with regard to phenomena such as meaning spectrum, collocations, antonymy and hyperonymy. Overall, this paper is a critical account of the current German lexicographic developments. It will include discussions on meta-lexicographic demands and focus on whether there are suitable complementary corpus approaches providing authentic dictionary information to a satisfactory extent.
|
