A Study Of The Lexicographic Treatment Of Academic Lexical Bundles In Advanced English Pedagogical Dictionaries

There is a tension between the word-centered tradition in practical lexicography and the phraseological tendency of language as revealed by corpus linguistic research.On the one hand,dictionaries,paper and electronic,have commonly treated the word as the primary indexical unit(i.e.headword)in the organization of entries.On the other hand,corpus studies have found that phraseologies often carry important pragmatic and discourse functions,and that some phraseological units are meaning bearers per se.Yet,the lexicographic treatment of phraseologies is usually less salient and adequate than that of individual words.At the same time,there has been a growing interest in developing English for academic purposes(EAP)phraseology lists to help non-native and novice writers with their academic writing,and lexical bundles—a distinct type of phraseology—have just begun to be accounted for in some online EAP dictionaries.It should be noted,however,that these lists were primarily built for pedagogical purposes and cannot be directly used as entry headwords,and that there is plenty of room for improving the description and presentation of lexical bundles in online dictionaries.In view of these gaps,this study set out to create an empirically derived,lexicographically useful list of lexical bundles which are characteristic of general academic writing and are deemed eligible for headword status.We used the Sketch Engine to extract an initial bundle list and went on to refine the list with replicable manual filtering and statistical measures,obtaining a total of 80 four-word target bundles.Then,we quantitatively analyzed whether and to what extent target bundles are treated in advanced English pedagogical dictionaries along the dimensions of bundle coverage,macro-and microstructures.In addition,in order to further improve bundle treatment in dictionaries,we compared two methods—the frequency approach and the collocation approach—for furthering bundle description,and found that the latter approach is more robust and effective than the former approach in extending the description of four-word bundles.Based on these findings,we proposed a new model for improving the lexicographic treatment of academic lexical bundles.The findings of this study can be summarized as follows.First,as regards methodology,this study demonstrates that it is feasible to derive a list of lexical bundles,which hopefully can be utilized in online English pedagogical dictionaries targeting non-native and novice writers.Second,it was found that the lexicographic treatment of target bundles varies by dictionary and is generally inadequate,based on the results of a quantitative analysis.Third,this study uncovers and empirically tests a new method(i.e.the collocation approach)for furthering bundle description,which has been underexplored in prior studies.This study has a number of implications.First,by focusing on the lexicographic treatment of corpus-derived phraseologies,this study makes a strong claim for granting headword status to a refined list of lexical bundles in online dictionaries with the aim of better satisfying non-native and novice writers‘ academic phraseological needs.Second,this study contributes to a synergy between lexical bundle research and electronic/digital lexicography.Third,this study showcases the potential of furthering the description of lexical bundles,and this may shed light on the construction of the ?ultimate dictionary‘(Sinclair et al.2004).A number of limitations and avenues for future research are also identified.First,this study focused on four-word bundles and did not take into account longer or shorter bundles that may also be frequently used in general academic writing.Second,currently the Oxford Corpus of Academic English can only be accessed via the Sketch Engine and as a result the bundle extraction process is subject to the Sketch Engine‘s N-grams settings,which more or less can influence the research result.Third,this study did not identify lexical bundles that are problematic to language learners.Fourth,this study did not examine the effect of different microstructural treatment on the look-up,retention and production of target bundles.It is hoped that future research will address these limitations.