LINGUIST List 32.1229
Wed Apr 07 2021
Review: Cognitive Science; Linguistic Theories: Nacey, Dorst, Krennmayr, Reijnierse (2019)
Editor for this issue: Jeremy Coburn <jecoburnlinguistlist.org>
Nina Julich-Warpakowski <nina_julich
Metaphor Identification in Multiple Languages E-mail this message to a friend Discuss this message
Book announced at https://linguistlist.org/issues/31/31-116.html
EDITOR: Susan Nacey
EDITOR: Aletta G. Dorst
EDITOR: Tina Krennmayr
EDITOR: W. Gudrun Reijnierse
TITLE: Metaphor Identification in Multiple Languages
SUBTITLE: MIPVU around the world
SERIES TITLE: Converging Evidence in Language and Communication Research 22
PUBLISHER: John Benjamins
REVIEWER: Nina Julich-Warpakowski, Universität Leipzig
“Metaphor Identification in Multiple Languages. MIPVU around the world” is an illustration and discussion of MIPVU ‘Metaphor Identification Procedure developed at Vrije Universiteit Amsterdam’, a method for reliable and replicable metaphor identification in discourse originally designed for English language data, and the problems and challenges that emerge when applying MIPVU to other languages. The edited volume begins with an introduction of the purpose of the book and a brief overview of the individual contributions. Chapter 2 is a reprint of the original MIPVU procedure from Steen et al. (2010). Chapter 3 presents additional guidelines for practically using MIPVU. Chapters 4 to 14 are the main body of the volume, each of which includes an application of MIPVU to a languages other than English (French, Dutch, German, Scandinavian languages, Lithuanian, Polish, Serbian, Uzbek, Chinese, Sesotho, and English as a lingua franca). Each chapter focuses on “operational issues” (e.g. available dictionaries, demarcation of linguistic units) as well as on more language-specific challenges (e.g. inflectional and derivational morphology). The book ends with an afterword by eminent metaphor scholar Elena Semino. In the following, each individual chapter will be briefly summarised.
Chapter 1, MIPVU in multiple languages (pp. 1-21), is an introduction to the book by the editors Susan Nacey, Aletta G. Dorst, Tina Krennmayr, and W. Gudrun Reijnierse, as well as eminent metaphor scholar and initiator of the development of MIP and MIPVU, Gerard J. Steen. While MIPVU as well as previous publications with a methodological focus on metaphor analysis in discourse (Cameron & Maslen 2010, Charteris-Black 2004, Heywood et al. 2002, Goatly 1997) have largely focused on English, metaphor identification procedures in languages other than English remain largely underrepresented. The present volume aims to fill this gap. Applications of MIPVU have started to appear (such as Badryzlova et al. 2013 for Russian) but the present volume presents a first and systematic collection of languages across the globe. This introductory chapter by Nacey et al. provides background on MIPVU and its predecessor MIP (Pragglejaz Group 2007). Other methods for metaphor identification are briefly discussed (Kittay 1984, Charteris-Black 2004, Cameron 2003). The chapter includes an interview with Gerard J. Steen providing insights into developing MIP and MIPVU.
Chapter 2, MIPVU: A manual for identifying metaphor-related words (pp. 23-40) by Gerard J. Steen, Aletta G. Dorst, J. Berenike Herrmann, Anna A. Kaal, Tina Krennmayr, and Trintje Pasma is a reprint of the original procedure from Steen et al. 2010. The procedure consists of a set of instructions to identify metaphor-related words in discourse. The notion ‘metaphor-related word’ is used to indicate that all lexical units identified as metaphorical by the procedure “can be taken to be lexical expressions of underlying cross-domain mappings” (p. 23). Details for each step are discussed in designated subsections. The basic unit of analysis in MIPVU is the orthographic word, or “lexical unit”. Metaphorical use of lexical units is identified by the following steps: 1) determining the contextual meaning of a lexical unit in the data with the help a dictionary (here the Macmillan English dictionary for advanced learners and the Longman dictionary of contemporary English), 2) determining whether the lexical unit has a more basic sense, i.e. a “more concrete, specific, and human oriented” sense in contemporary language use (p. 32), 3) determining whether both basic and contextual meaning are sufficiently distinct, and 4) deciding on whether both meaning can be related by some form of similarity.
In Chapter 3, What the MIPVU protocol doesn’t tell you (even though it mostly does) (pp. 41-67) by Susan Nacey, Tina Krennmayr, Aletta G. Dorst, and W. Gudrun Reijnierse, potential challenges applying MIPVU are illustrated and further practical guidelines are suggested. The authors focus on aspects of the “nitty-gritty” details (p. 41) of the procedure that novice users often tend to overlook. These regard deciding on what counts as a lexical unit, determining a more basic meaning of a lexical unit, and establishing whether contextual and basic meaning constitute a (metaphorical) comparison. Nacey et al. also discuss and illustrate the use of software tools, such as Excel (for which a MIPVU annotation template is provided in the volume’s online supplementary material), File Maker, or R (R core team 2013). The chapter concludes a brief illustration of the usefulness and rationale of inter-rater reliability testing. Nacey et al. also provide more general advice on the application of MIPVU to a research topic.
In Chapter 4, Linguistic metaphor identification in French (pp. 69-90), W. Gudrun Reijnierse suggests adjusted guidelines for the identification of metaphors based on language-specific lexical and grammatical characteristics of French. The chapter includes an illustration of the procedure applied to one sentence from the French newspaper Le Figaro. She presents a study, in which the same corpus of approximately 45,000 words was analysed twice, each time using a different dictionary. The dictionaries used were the electronic versions of Le Petit Robert, and Le Grand Robert & Collins. The results reveal that for 96.86% of the lexical units in the corpus both dictionaries yield the same result. There are however more considerable differences when it comes to the (non-)metaphorical status of lexical units with respect to word classes: identification particularly deviates for adverbs, nouns, and verbs. The author discusses qualitative differences in the organisation of entries in each dictionary that may account for these discrepancies.
Chapter 5, Linguistic metaphor identification in Dutch (pp. 91-112) by Tryntje Pasma, is a reprint of Chapter 7 in Steen et al. 2010 and reports the author’s experience with the application of MIPVU to a corpus of Dutch. Since no corpus-based dictionary for Dutch exists, the historically based Van dale dictionary was used. Pasma shows that the three analysts who coded the data agreed on the metaphorical status of more than 90% of the words analysed, which indicates that MIPVU can be well adopted to Dutch, bearing in mind minor language-specific adjustments. One issue concerns separable complex Dutch verbs, which can be separated into verbal root and particle. In analogy to English phrasal verbs, these are always considered as one unit. Another issue concerns polywords such as “met name”, ‘in particular’. These are not listed in the dictionary and thus have to be analysed into their component units, which often poses a problem since the metaphorical status of the individual components is difficult to establish.
In Chapter 6, Linguistic metaphor identification in German (pp. 113-135), J. Berenike Herrmann, Karola Woll, and Aletta G. Dorst show how MIPVU can be applied and adjusted to identify metaphors in German data based on the analysis of a sample of 20,549 words from different genres. The data was automatically POS-tagged. Due to language-specific characteristics of German and the nature of the dictionaries used, the following issues call for an adjustment of MIPVU to German: the demarcation of lexical units (especially with respect to separable complex verbs (like Dutch above), compounds, and polywords), non-autonomous sense descriptions in the dictionary (e.g. a nominalisation derived from a verb may not have a separate entry), and establishing distinctness between senses. Each issue is discussed in detail and a set of clearly stated guidelines is provided. The authors report reliability results for the analysis of a subset of the data. Reliability for metaphorically used words is moderate (Fleiss’ Kappa: 0.71) suggesting that MIPVU can be well applied to German.
In Chapter 7, Linguistic metaphor identification in Scandinavian (pp. 137-158), Susan Nacey, Linda Greve, and Marlene Johansson Falck present a metaphor identification procedure for Danish, Norwegian, and Swedish. As in the previous chapters, adjustments have to be made when applying MIPVU to Scandinavian languages. The demarcation of lexical units should correspond to the orthographic word, i.e. complex words written solidly should be considered one unit (reflexive verbs are discussed as an exception). As regards polywords, the authors have created their own preliminary lists for each of the three languages in part based on the Oslo-Bergen tagger. The procedure for metaphor identification in Scandinavian is presented in a nutshell, and illustrate in a sample analysis. The authors report reliability of a case study, which was analysed in two rounds. For identification of metaphor-related words, initially there was only moderate agreement, which was largely due to the fact that the metaphoricity of prepositions was overlooked or unclear, but strong to almost perfect agreement after round two.
In Chapter 8, Linguistic metaphor identification in Lithuanian (pp. 159-181), Justina Urbonaitė, Inesa Šeškauskienė, and Jurga Cibulskienė present an application of MIPVU to Lithuanian. In the absence of a contemporary corpus-based dictionary of Lithuanian, the use of a number of lexical resources is suggested. The authors discuss language-specific challenges that call for an adjustment of MIPVU when applied to Lithuanian. Particularly this concerns the rich inflectional nature of Lithuanian. The authors consider abstract nouns that are marked for dative, locative, or instrumental case as metaphorically used thus deviating from the original MIPVU by considering a lexical unit’s internal morphological structure. The same applies to derivational affixes, which often have prepositional meanings. The authors present an exemplary application of MIPVU to selected lexical units. In a case study, three analysis applied the Lithuanian MIPVU to 1,305 words of academic legal discourse scoring an impressive Fleiss’ Kappa of 0.94, which is considered “almost perfect agreement”.
In Chapter 9, Linguistic metaphor identification in Polish (pp. 183-202), Joanna Marhula and Maciej Rosiński present an application of MIPVU to Polish language data. Like Lithuanian, Polish is a language rich in inflections and derivational morphology which calls for adjustments of the procedure. Adjustments are based on the exhaustive analysis of a dataset of 7,604 words by two analysts. The authors discuss the lexical unit status of multiword expressions, reflexive verbs, prepositional verbs, as well as proper names. As regards case inflections, Marhula and Rosiński follow the suggestion of the authors of the previous chapter and consider case endings to be potentially metaphorical. The authors illustrate this with an example of the vocative form of inanimate nouns, which can be interpreted as a personification. Finally, an illustration of the adjusted procedure is presented. As regards reliability of the procedure, the authors report a Cohen’s Kappa value of 0.77 for their analysis.
In Chapter 10, Linguistic metaphor identification in Serbian (pp. 203-226), Ksenija Bogetić, Andrijana Broćić, and Katarina Rasulić present an application of MIPVU to Serbian, which is also intended to be compatible to other standard varieties of Serbo-Croatian, i.e. Croatian, Bosnian, and Montenegrin. The main part of the chapter focuses on the Serbian case system. In order to account for this “hidden” case-encoded metaphoricity, the authors suggest including an additional step for the application of MIPVU to Serbian, in which the analyst should establish the contextual meaning for each case inflection and determine whether it has a more basic meaning. For that purpose, the authors provide a list of potential basic senses for the genitive, dative, accusative, and instrumental case (p. 205). In a case study, the authors illustrate the applicability of the procedure. Reliability scores are comparatively high, both for lexical metaphor identification (Cohen’s kappa = 0.93) as well as “inflectional metaphor” identification (Cohen’s kappa = 0.83).
In Chapter 11, Linguistic metaphor identification in Uzbek (pp. 227-245), Sıla Gen Kaya presents an application of MIP, a predecessor to MIPVU, to the Turkic, agglutinating language of Uzbek. The paper presents suggestions for adjustments of MIP to the study of metaphor in Uzbek texts based on a sample of 80,000 words of contemporary Uzbek fiction (Gen 2015). Gen discusses issues with respect to the use of a dictionary. Regarding linguistic issues, she discusses the potential metaphoricity of inflectional case endings of the locative, dative and ablative case, and in line with previous chapters suggests that abstract nouns with such a case ending should be regarded as metaphorically used given the fact that case endings have a more basic meaning referring to spatial relations. No reliability results are reported since there was only one analyst. Gen Kaya still affirms that “when meticulously applied, MIP [as well as MIPVU] provides valid results in Uzbek discourse” (S. 244).
In Chapter 12, Linguistic metaphor identification in Chinese (pp. 247-265), Ben Pin-Yun Wang, Xiaofei Lu, Chan-Chia Hsu, Eric Po-Chung Lin, and Haiyang Ai present an application of MIPVU to Chinese. The corpus material annotated for this purpose is sampled from the Lancaster Corpus of Mandarin Chinese (LCMC). Given that the Chinese orthographic system is character based, the fact that there is no clear word delimiter poses a problem for lexical unit demarcation. The data sampled from the LCMC, however, provides automatic word segmentation. Chinese-specific types of lexical units such as verb-object compounds, resultative verb compounds, and reduplications as well as the handling of proper names, idioms, and fixed expressions are discussed. Despite some challenges, the authors affirm that an application of MIPVU to Chinese data especially in a processed and prepared format as provided by the LCMC is “largely smooth” (p. 263) as is illustrated in a sample analysis. Reliability results are not reported (but see Lu & Wang 2017).
In Chapter 13, Linguistic metaphor identification in Sesotho (pp. 267-287), Nts’oeu Seepheephe, Beatrice Ekanjume-Ilongo, and Motlalepula Raphael Thuube present an application of MIPVU to the agglutinating Bantu language of Sesotho spoken in Lesotho and parts of the Republic of South Africa. The authors discuss dictionary use as well as language-specific issues of Sesotho which call for some adjustments when applying MIPVU, such as the demarcation of lexical units, verb extensions, reduplications, and markers for the diminutive and the augmentative. Seepheephe et al. then present an adjusted protocol of MIPVU for metaphor identification in Sesotho and demonstrate its application through an illustrative sample analysis of selected lexical units from the novel Chaka by Thomas Mofolo. For the present study, reliability of the procedure was not measured, and is suggested as a desideratum for future research.
In Chapter 14, Linguistic metaphor identification in English as a lingua franca (pp. 289-312), Fiona MacArthur discusses how ELF characteristics may pose challenges to metaphor identification with MIPVU. She presents a detailed description of her data, the EuroCoAT corpus (size = 55,718 words) and its heterogenous nature comprising data from speakers with English as their first (L1), second (L2) or third language (L3). She suggests, for example, that polywords can be regarded as decomposable if there is potential for variation (e.g. “a (little) bit of”) given that these might be less entrenched as chunks for non-L1 speakers of English. In that light, MacArthur also discusses the potential fixedness/decomposability of phrasal verbs and compounds as well as the fact that dictionaries do not necessarily mirror non-native speaker competence of English. The chapter closes with a brief illustration of the modifications suggested for MIPVU when applied to ELF data.
The final chapter of the volume, Chapter 15, presents an “Afterword: Some reflections on MIPVU across languages” by metaphor scholar Elena Semino (pp. 313-321). Semino emphasises how “MIP and MIPVU have dramatically changed the field of metaphor research for the better” (p. 315) by rendering metaphor analyses reliable, replicable and comparable. She stresses that the volume by Nacey et al. is so important because it not only includes European languages from different language families but takes on a much more international perspective. She, however, also reminds scholars that MIPVU is a means to an end, and that it should be applied flexibly in line with the overall research goals. Moreover, although MIPVU uses a binary distinction of metaphor — either a word is metaphorically used or not — the phenomenon in reality is often multi-faceted, scalar, and multimodal (p. 321, see also Müller 2008, Müller & Tag 2010, Jensen 2017).
While the final chapter by Elena Semino already presents an excellent summary and evaluation of the book, some additional aspects regarding the quality of “Metaphor Identification in Multiple Languages. MIPVU around the world” edited by Nacey, Dorst, Krennmayr, and Reijnierse shall be briefly provided in the following paragraphs.
The edited volume is an excellent resource for researchers who want to apply the procedure, and adopt it to a specific language, either one of the languages discussed in the book, or one that has not been dealt with so far. Furthermore, it is a good resource for advanced MIPVU users, researchers and university lecturers who would like to teach the procedure in English or any other language. Here, especially Chapter 3 presents a welcome addition to the original procedure in Chapter 2 since it spells out common pitfalls and areas in which MIPVU is best applied. This is done in a very approachable manner and based on the editors extensive practical experience in using and teaching MIPVU.
The book is very well edited. All chapters are similarly structured, address similar issues, and also label these issues similarly, which renders the volume a very coherent whole. Apart from that, each chapter is written in a self-contained manner and thus readers interested in the application of MIPVU to one specific language, may only consult that chapter. Given that the volume includes a reprint of the original MIPVU procedure, the edited volume itself is self-contained and presents a good resource for students and researchers alike, ready to get their hands dirty and to analyse metaphor in the wild across the globe. Working with the book is also facilitated by the presence of an index of the most relevant terms and concepts. Finally, one of the main advantages of this book is that it comes with an online repository (https://osf.io/vw46k
) where readers get access to templates for carrying out there own analyses, individual languages’ polyword lists, and R scripts to perform reliability tests. Through this supplementary, the authors keep in line with the goal of making metaphor analysis transparent, and encourage researchers to carry out their own analyses.
The book may thus inspire methodological discussion of metaphor research, as well as future application to more languages. As the editors state in their introduction: “With over 6,000 languages worldwide, this volume clearly will not be the final word on the topic, although it does represent a solid start” (p. 18).
Badryzlova, Y., Isaeva, Y., Shekhtman, N., & Kerimov, R. (2013). Annotating a Russian corpus of conceptual metaphor: A bottom-up approach. Proceedings of the Workshop on Metaphor in NLP, 77–86. http://aclweb.org/anthology/W13-09
Cameron, L. (2003). Metaphor in educational discourse (Vol. 3). continuum.
Cameron, L., & Maslen, R. (2010). Identifying metaphor in discourse data. In L. Cameron & R. Maslen (Eds.), Metaphor Analysis. Research Practive in Applied Linguistics, Social Sciences and the Humanities (pp. 97–115). Equinox.
Charteris-Black, J. (2004). Corpus approaches to critical metaphor analysis. Palgrave Macmillan.
Gen, S. (2015). Özbekçede metaforlar [Metaphors in Uzbek] [Doctoral dissertation]. Cukurova University.
Goatly, A. (1997). The Language of Metaphors. Routledge.
Heywood, J., Semino, E., & Short, M. (2002). Linguistic metaphor identification in two extracts from novels. Language and Literature, 11, 35–54.
Jensen, T. W. (2017). Doing Metaphor: An Ecological Perspective on Metaphoricity in Discourse. In B. Hampe (Ed.), Metaphor: Embodied Cognition and Discourse (pp. 257–276). Cambridge University Press.
Kittay, E. F. (1984). The identification of metaphor. Synthese, 58(2), 153–202.
Lu, X., & Wang, B. P.-Y. (2017). Towards a metaphor-annotated corpus of Mandarin Chinese. Language Resources & Evaluation, 51(3), 663–694.
Müller, C. (2008). Metaphors Dead and Alive, Sleeping and Waking. A Dynamic View. The University of Chicago Press.
Müller, C., & Tag, S. (2010). The Dynamics of Metaphor: Foregrounding and Activating Metaphoricity in Conversational Interaction. Cognitive Semiotics, 10(6), 85–120.
Pragglejaz Group. (2007). MIP: A method for identifying metaphorically used words in discourse. Metaphor and Symbol, 22(1), 1–39.
R Core Team. (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing. http://www.R-project.org
Steen, G. J., Dorst, A. G., Herrmann, J. B., Kaal, A. A., Krennmayr, T., & Pasma, T. (2010). A method for linguistic metaphor identification: From MIP to MIPVU. John Benjamins.
ABOUT THE REVIEWER
Nina Julich-Warpakowski holds a PhD in English linguistics. Her focus is on conceptual metaphor, empirical approaches to metaphor, the gradable nature of metaphor, as well as the relation of metaphor to other figurative processes such as fictive motion. She works at Leipzig University, Germany, as a researcher and teacher of English linguistics.
Page Updated: 07-Apr-2021