INTRODUCTION TO THE CORPUS ANALYSIS OF PHRASEOLOGY

Аннотация

This article argues that the use of digitized dictionary texts for corpus analysis offers numerous advantages and opens up expanded perspectives in phraseology. The systematic use of such resources can provide deeper insights into the development and structure of phrasemes, which is invaluablefor linguisticresearch. Keywords:Phraseology, corpus analysis, phrasemes, idiomsand proverb lexicon, corpora. A successful corpus analysis in the field of (historical) phraseology requires input that is independent of expert intuition. The goal of this analysis is to search large corpora in a targeted manner to find meaningful evidence for phrasemes[3]. These evidences are crucial for a) establishing a chronological dating and b) analyzing, verifying, and describing the semantic structureof thephraseme in context. The procedure is similar to that of single-word lexicography, with the addition of a search for variants in phraseography. A central problem is the lack of specific input necessary to know which phrasemes and their variants existed historically. An example illustrates this problem: The phrase "jemandem die Leviten lesen" (to read someone's riot act) is considered common today, butnolexical variants areknown. The analysis of the DWDS core corpus shows that "Leviten" appears as a collocator only 139 times, without identifying any further variants.[5] In addition, the historical German Dictionary of Proverbs provides at least some lexical variants, such as "Epistel" or "Vers." However, it remains unclear how common these variants actually were, making cross-checking with a text corpus essential. Corpus analysis versus looking up dictionaries. Looking up the Duden dictionary of idiomatic expressions doesn't yield many new insights, as it lacks further references. The German Dictionary of Proverbs couldn't help much either, as the high hit rate is primarily due to the explanations ofthe common variant. However, thefull-text search in thedigitized German Dictionary of Proverbs makes it possibleto discover both lexical variants and phraseological synonyms. For example, alternatives such as "Ich hab' ihmden Kümmelgerieben" (Irubbed himthecaraway) can befound. Digital corpus analysis. This approach is far more than simple reference; it is corpus research. Digital access to full texts transforms printed reference works into electronic corpora, opening up newpossibilities for phraseological research. Corpus analysis plays a central role in phraseological research. This particularly concerns the search for phrasemes, which is often treated as a special case. According to Rothkegel, it is about https://ijmri.de/index.php/jmsi volume 4,issue4,2025 722 determining under which conditions a word chain is considered a lexical-semantic unit and how this can be determined. The most commonly used method for identifying multi-word units is collocation analysis, which analyzes large numbers of neighboring lexemes in a short time and thusprovides valuableclues to phraseologized word combinations. Despite these advances, Heid still sees a need for development, especially with longer and more varied phrases. Many idiomatic combinations are only recognized automatically, which is effective, but does not capture all facets of the language. The interpretation of whether a phraseme is a phrase is often left to experts. These must not only evaluate the semantic features, but also determine the variants and boundaries of the phrasemes, which means additional research effort. A particular challenge is the historical dimension of phrasemes, which arises from their cultural embedding and motivational history.[2] This leads to a competence problem, as experts cannot easily assess how phrasemes functioned at different historical language stages. The availability of contemporary sources decreases with time as we look back through language history, further complicatinganalysis. The automatic extraction of phraseological material, especially for more complex phrases, proves unrealistic. Collocation analysis is highly dependent on sufficient data; for earlier centuries, such as the 19th century, data are often sparse. Statistical conclusions are therefore often unreliable, as relevant evidence formanyvariants is missingor scarce. Overall, it appears that the focus on idiomatic phrasemes and a diachronic perspective significantlyworsens thecost-benefit ratio ofcorpus analysis. Search for linguistically related noun forms. In (historical) phraseology, real evidence is of central importance, as it provides information about the variants and morphosyntactic variations in which phrasemes existed[5]. Verifying the actual occurrence of these phrasemes in large historical corpora is therefore essential. However, a lack of evidence does not negate the fundamental existence of aphraseme. Evidential dictionaries play a crucial role here, as they list realia for individual lemmas and phrasemes and thus confirm the actual occurrence of the lemma. While the "Gram