Advances in Corpus-based Contrastive Linguistics (henceforth ACCL) is a collection of 12 papers on corpus-based contrastive studies in honor of Professor Stig Johansson, a pioneer in corpus linguistics and in corpus-based contrastive research.Footnote [1] As pointed out in the editors’ introduction, Johansson and his team’s creation of the English–Norwegian Parallel Corpus (1997) and its later development into the Oslo Multilingual Corpus (2008) made it possible to carry out contrastive analysis based on naturally occurring data rather than researchers’ intuitions. According to the editors of this volume, the corpus-based approach has ‘initiated a new era in contrastive linguistics’ and ‘placed contrastive analysis on a sound empirical footing’ (1). This review will first present the contributions one by one, and then provide an evaluation of the related methodological and theoretical issues in the wider context of contrastive linguistics.
In the first chapter, entitled ‘Tertia comparationis in multilingual corpora’, Thomas Egan employs the original Norwegian expression mellom ‘between’ as the Tertium Comparationis (TC; see Krzeszowski Reference Krzeszowski1990: 15; Johansson Reference Johansson2007: 39) for encoding the concept of ‘betweenness’, and makes a comparison between its translation equivalents in English and French. The study establishes that there is a considerable overlap between English and French in their use of default prepositions between and entre while French employs non-prepositional constructions to a greater extent. It also argues that taking a third language as TC and comparing its translations in two other languages could bypass the question of ‘what exactly is being compared’ and ‘stipulate the degree of resemblance between the two languages’ (23).
The second chapter is ‘Seeing the lexical profile of Swedish through multilingual corpora’, by Åke Viberg. He argues that ‘vehicle verbs’ are typologically significant as a sub-type of motion verbs. Drawing on the self-built multilingual parallel corpus, Viberg makes a comparison between three Swedish verbs describing motion in a vehicle, åka ‘go’, fara ‘go’ and köra ‘drive’, and their translation equivalents in English, German, French and Finnish. It turns out that the semantic extensions of individual verbs differ across even closely related languages. In addition, the inventory of general vehicle verbs and ways of coding a certain type of situation are language-specific. Thus, Talmy’s (Reference Talmy2000) dichotomy of verb-framed and satellite-framed languages should be regarded ‘as a continuum’ (25).
The next chapter is ‘English affixal negation translated into Spanish’, by Rosa Rabadán & Marlén Izquierdo. Regarding translation equivalence as TC, the authors’ focus is the English negative affixes and their Spanish translations as found in a parallel corpus. By using a monolingual corpus as control data, they show that the translated usage of English negative affixes deviates significantly from their non-translated usage. The translation texts are more like the source language rather than the target language, which differs from previous findings that translators have a tendency to conform to expressions that are typical of the target language. However, as the authors point out, replication of the tests using other corpora is required in order to validate the findings.
In the next chapter, Anne-Marie Simon-Vandenbergen explores ‘English adverbs of essence and their equivalents in Dutch and French’ from a translation perspective, testing Christopher Butler’s (e.g. Butler Reference Butler2008) conclusion based on an investigation of monolingual corpora. According to Butler (Reference Butler2008), the English adverbs basically, fundamentally and essentially differ in terms of their relative frequency, semantic preference and syntactic behavior. In this study, Simon-Vandenbergen first examines the semantic-pragmatic behavior of these adverbs in the monolingual British National Corpus (BNC) and then analyzes their French and Dutch equivalents drawing on a bidirectional trilingual corpus. This study of cross-linguistic equivalents corroborates Butler’s findings and throws more light on the three words.
In ‘A parallel corpus approach to investigating semantic change’, Kate Beeching explores the semantic evolution of French pragmatic marker quand même ‘even so’ on the basis of diachronic parallel corpora by treating translation equivalence as the observed TC. This chapter contributes to a long-standing debate on the meaning of pragmatic markers (see e.g. Hansen Reference Hansen1998, Aijmer Reference Aijmer2002): should a particular lexeme be considered as having a ‘core meaning’ and contextual side effects or should each pragmatic function of a lexeme be considered a separate sense, thus rendering the lexeme polysemous? The study reveals that quand même is synchronically polysemous with its core meaning almost entirely bleached, and that it is multifunctional, serving either adversative or relational functions. The translators’ systematic selection of a particular functional equivalent proves to be a sound test for semantic change. The author also argues that translation is a useful means of demonstrating that semantic change has indeed occurred.
In the next chapter, ‘Youngspeak: Spanish vale and English okay’, Anna-Brita Stenström investigates the use of pragmatic markers in young people’s language, focusing on the way they function on the interactional, interpersonal and discourse levels of conversation. Based on two large comparable corpora, the study finds that although vale is slightly more frequent in Spanish than okay in English, they are both multifunctional and most of their functions converge. A further comparison in terms of their frequency from a sociolinguistic point of view reveals that the two pragmatic markers are used more by teenagers than adults.
Sylvie de Cock & Diane Goossens research ‘Quantity approximation in English and French’, focusing on the approximators that ‘appear to be largely unexplored in contrastive studies’ (141). This study combines the use of comparable corpora with an ‘inductive corpus-driven approach’ to extract linguistic devices indicating quantity approximation (e.g. AROUND 1350 jobs, AT LEAST five years). The study demonstrates that approximators in both languages exhibit similar semantic tendencies, including those expressing a certain quantity, a minimum quantity, a maximum quantity and an interval quantity. The collocates of the approximators are also similar in English and French. In terms of the grammatical categories of the approximators, only the top three (adverbs, more/less than, and prepositions) are shared by the two languages. Overall, English tends to use imprecision around numbers to a greater extent.
In ‘Enriching the phraseological coverage of high-frequency adverbs in English–French bilingual dictionaries’, Sylviane Granger & Marie-Aude Lefer investigate the application of monolingual as well as translation corpora in bilingual lexicography. Taking English yet and French encore ‘still’ as their focus, the authors compare their dictionary entries with corpus data, paying special attention to the coverage and presentation of the lexical bundles including the two words. They recommend increasing the number of phrases, the accuracy of their translation equivalents and the authenticity of the examples in bilingual dictionaries. They also argue that using corpora makes it possible to avoid categorical bias in translations.
In the next chapter, entitled ‘Using recurrent word-combinations to explore cross-linguistic differences’, Jarle Ebeling, Signe Oksefjell Ebeling & Hilde Hasselgård explore phraseological differences between English and Norwegian. Three case studies are selected from the three-word combination list, i.e. the construction of the form of ‘PREP det ADJ’ (e.g. i det minste ‘in the least’), all the way/all the same, and three-word time expressions (e.g. for a moment). The contrast is made on the basis of a bidirectional parallel corpus, and points to constructional, semantic and pragmatic differences in the two languages (177). It also turns out that similar expressions in two languages (e.g. all the way in English and hele veien ‘all the way’ in Norwegian) may be ‘associated with different degrees of metaphorical potential and thereby different conditions of use’, and that they need not occur ‘with the same frequency in comparable texts’ (198).
The remaining three chapters focus on discourse. Kerstin Kunz & Erich Steiner present a contrastive account of ‘Cohesive substitution in English and German’. They start by outlining the relationship between the three types of cohesion, i.e. reference, substitution and lexical cohesion, pointing out that the function/semantics of substitution is distinct from that of the other two, and that its ‘lexico-grammatical realizations’ are specific to its type. Treating the semantic/function level of the substitution as TC, they analyze the systemic differences between German and English with respect to their nominal, verbal and clausal substitution mechanisms. The study reveals that substitution is more common in English than in German whereas German uses a greater variety of forms.
Next, Jennifer Herriman carries out a study on ‘The extraposition of clausal subjects in English and Swedish’. Sentences containing that-clauses and infinitival clauses in the two languages are extracted from the parallel corpus, after which the author compares the extent to which those embedded clauses representing the subject of the sentences are extraposed. The study shows a more frequent use of extraposition in Swedish. The authors also indicate that ‘subject clauses are more frequent in Swedish, both when they are extraposed and when they are retained in preverbal position’ (233).
The final chapter is ‘Thematic variation in English and Spanish newspaper genres’, by Julia Lavid, Jorge Arús & Lara Moratón. The authors concentrate on the clausal and discourse thematic features of two newspaper genres, i.e. news reports and commentaries, as well as the genre-specific or language-specific constraints on the selection of those features. At the clausal level, mainly the Thematic Head, defined as ‘the first nuclear experiential constituent within the main clause’ (266), is analyzed. In terms of the discourse thematic features, on the other hand, the authors look at three aspects of the Theme: how different Themes recognize different moves characterizing the two genres, the patterns of thematic development and the inner structure of the Themes. This study indicates a stronger relationship between genre and thematic selection, although language-specific preferences also play a role in the observed variation.
ACCL makes a valuable contribution to contrastive linguistics, with obvious advances in its exploitation of corpora and corpus-driven techniques to a wide range of topics. A variety of corpora are discussed in this volume, including bi/multilingual parallel corpora and monolingual comparable corpora. Corpus-linguistic methodology like part-of-speech tagging, extraction technique and recurrent word combinations is applied to provide points of departure for the comparison instead of pre-selected linguistic expressions. Such methodological approaches have provided bottom–up foundations for contrastive studies, making the comparisons objective and empirically sound. The topics under investigation are not limited to comparisons of specific lexical categories, collocations or syntactic constructions, but explore ‘new domains such as semantic change, pragmatic markers and discourse phenomena’ (3). Although semantic and pragmatic features in discourse are typically ‘not traceable at the formal level’ (Marzo, Heylen & de Sutter Reference Marzo, Heylen, de Sutter, Marzo, Heylen and de Sutter2012: 2), a combination of qualitative analysis with the corpus-based approach has proved to be reliable, as reflected in the studies on discourse phenomena in the last three chapters.
TertiumComparationis, the ‘common platform of reference against which differences can be stated’ (Krzeszowski Reference Krzeszowski1990: 15), has always been a controversial notion in contrastive studies. ACCL, however, witnesses its development from a new perspective with the wide application of corpora. As James (Reference James1980: 178) puts it, ‘translation equivalence is the best available basis of comparison’. Parallel corpora can thus serve to provide systematic evidence, as shown in Egan’s study of ‘betweenness’ in English and French, with the Norwegian mellom taken as TC. This approach helps ‘stipulate the degree of resemblance between the two languages in a way [that approaches drawing on just the two languages] cannot’ (23). Monolingual comparable corpora, on the other hand, can also provide reliable basis for cross-linguistic studies in spite of the drawback of ‘lacking an obvious TC’ and the question of ‘not knowing what to compare’ (3). Indeed, we can have a clear ‘background of sameness against which differences are viewed’ (Johansson Reference Johansson2007: 39) based on theoretical linguistics findings in different languages. The semantic/functional features of cohesive substitution in English and German respectively, for instance, serve as TC in Kunz & Steiner’s study of cohesive substitution. Thus, TC, as a common platform for all comparisons, still proves to be an essential notion in contrastive studies, especially with the availability of corpus.
ACCL demonstrates clearly that empirical research does not stand in contrast with theoretical hypotheses. The corpus-based approach in ACCL differs from traditional corpus-linguistic approaches that confine analysis to the corpus data (e.g. collocations, wordlists, keywords). Instead, corpora in this volume serve as empirical evidence and allow researchers to set up hypotheses about cross-linguistic differences. Current trends have shown that contrastive studies are being carried out in various linguistic approaches, in particular from functional and cognitive perspectives (e.g. Gómez González, Mackenzie & González Álvarez Reference Gómez González, Lachlan Mackenzie and González Álvarez2008, Boas Reference Boas2010). These studies also reveal that the usage-based approach is a point of convergence among functional, cognitive and corpus linguistics. Just as Marzo et al. (Reference Marzo, Heylen, de Sutter, Marzo, Heylen and de Sutter2012: 2) rightly note, contrastive linguistic study in the future calls for ‘a corpus-driven approach and its interaction with theoretical findings’.
As the editors state in the introduction, ‘corpus-based contrastive analysis has experienced a dramatic development since the 1990s’ (3). The studies in ACCL demonstrate the potential of corpora in the comparison of languages at various levels from lexis to discourse, promising exciting new developments in this field. The volume thus proves itself to be a worthwhile read for contrastivists and corpus linguists, as well as typologists.
