مهدی باقری حریری

The purpose of the present study was to propose a practical and corpus-based solution for the problems of meaning discrimination and equivalent selection by Hartmann’s Contrastive Textology Theory (۱۹۸۰) and the use of parallel corpora primarily in the process of translation, and secondly, extending it to the compilation of bilingual dictionaries. It also aimed to suggest the most reliable way of finding more potential equivalents of lexical items and their different meanings in various contexts. To that end, the researcher used firstly Mizan Persian-English Parallel Corpus and secondly TEP (Tehran English-Persian Parallel Corpus) to prepare the required bilingual corpora. The two available corpora (Mizan and Tehran) were suitable for the research purposes respectively. Firstly, one hundred polysemous content words were chosen by the researcher from both corpora as search words that were to be selected based on frequency word lists provided by Wordsmith as the corpus analysis tool. Secondly, these words were searched in the English corpus and different Persian equivalents of them were found and extracted from the aligned sentences in the parallel corpora where various meanings and senses of the search words were shown. Thirdly, the found Persian equivalents were collected and listed in tables. Fourthly, the Persian translations and equivalents in the parallel corpora were compared with Persian equivalents in Farhang Moaser English-Persian Millennium Dictionary that is a famous bilingual English-Persian dictionary to find the different meanings and senses. Finally, the equivalents available in Glosbe online dictionary were provided as well for the purpose of comparison to see which source (parallel corpora, Glosbe or bilingual dictionary) provides translators with more comprehensive and detailed equivalents based on different linguistic contexts. The findings indicated that because of the fact that the available corpora are very general and small and the reviewed bilingual dictionary is not based on a large corpus, there were equivalents in both the parallel corpora and Glosbe which were absent in the dictionary and there were some meanings in the dictionary not found in the corpora as well. Generally, the results showed that parallel corpora can give us more potential equivalents and in addition to having some problems and weaknesses, the existing limited corpora do not contain the required variety of equivalents.