Lars Hellan, Andrej Malchukov, and Michela Gennamo (eds), Contrastive Studies in Verbal Valency

LARS HELLAN, ANDREJ MALCHUKOV AND MICHELA GENNAMO (EDS), CONTRASTIVE STUDIES IN VERBAL VALENCY LINGUISTIK AKTUELL/LINGUISTICS TODAY, VOLUME 237. Amsterdam/Philadelphia: John Benjamins PublishingCompany, 2017. ISBN 9789027257208 Ever since Tesnière (1959), valency has been acentral topic in syntax. The traditional work on valency shares two characteristics. First, it takes on alexical view of valency, i.e. alexical unit (typically averb) is said to have acertain valency frame, and valency alternations are explained in terms of lexical unit derivation. This view has not gone unchallenged; probably the best-known alternative was pioneered by Adele E.Goldberg in her now-classic monographs (1995, 2006). Second, linguists have typically studied valency in individual languages, and have usually focused only on selected facets of the complex topic, thus leaving agap that still remains to be fully bridged. Only recently the efforts to do that have given rise to various valency-centered enterprises; two examples that could be paid heed to are the Leipzig Valency Classes Project (cf. Malchukov and Comrie 2015) and the Construction Labeling project (cf. Hellan and Dakubu 2010). The monograph under review certainly comes as awelcome contribution to the field, bridging the gap alittle further, while also tackling some of the theoretical questions relating to the first above-mentioned characteristic. The authors of the studies included in the monograph are affiliated with various institutions, interested in diverse languages from all over the world, and their contributions are anchored within different linguistic frameworks. All of them came together, however, sharing the interest in valency of Mary Esther Kropp Dakubu (1938–2016), to whose memory the volume is dedicated. She was astudent of linguists so prominent as Zellig Harris; later she moved to Ghana, where she spent her whole professional life, conducting research on Niger-Congo languages, becoming an expert in various fields of West African linguistics. She left behind an impressive richness of results on which others can build. The thirteen studies included in the book are organized into three major parts, revolving around argument coding, valency rearranging alternations, and the interface between valency and grammatical voice, respectively. The papers are accompanied by abrief introductory overview of the notion of valency, which pays special attention to the descriptive strategies applicable when dealing with verbs that may occur in several valency frames. The introductory paper also considers the link between verb semantics and valency frames, illustrating how contrastive studies might facilitate the research in this area. The following survey is necessarily selective for lack of space; Ican only comment very briefly on each paper, and without trying to be exhaustive, Itake the liberty of adding an occasional remark that might sparkle the interest of others. The first four studies explore the area of valency frames and argument coding. The first paper, written by Werner Abraham and Elisabeth Leiss, is titled Multiple case binding— The principled underspecification of case exponency. The authors pursue Jakobson’s notion of paradigmatic case, arguing against syntactic case, i.e. case being viewed as syntagmatically structured (aview due to Tesnière). The authors Jakub Sláma 242 LINGUISTICA PRAGENSIA 2/2018 insist that there should be aturn towards the paradigmatics of case, and that the research in case functions should start anew by redefining the category adequately. They also take the opportunity to criticize both formalist approaches (for “areduction of case to its formal side”, p. 29) and Construction Grammar. As for the latter, they claim that “syntagmatic case as represented in case frames or verb frames is the favorite topic of construction grammarians to enforce their claim that syntax is nothing more than idiomatic construction” (p.74), dubbing this a“dubious” methodological approach. While any constructive criticism is welcome by the members of the cognitive-constructionist fraternity, Ibelieve that very few of them— if any— would consider syntax to be “nothing more than idiomatic construction.” Dorothee Beermann’s study Infinitives deals with control patterns in German and Norwegian bare infinitive constructions, focusing on aspectual and perception verbs. She relies on the framework of Lexical Functional Grammar, which allows her to pinpoint functional and semantic constraints underlying the different properties of German and Norwegian verb chains. Mary Esther Kropp Dakubu and Lars Hellan’s contribution titled Alabeling system for valency: Linguistic coverage and applications presents the Construction Labeling System, aformat for valency representation that has proved to be across-linguistically applicable tool for valency-related annotation, as illustrated with examples from English, Norwegian, and Ga (spoken in Ghana). Given the limitation of space, Ican mention only three points that Ifind worth noting. First, the authors work with the notion of constructions. While they make it clear that they consider their notation as “neutral relative to the framework ‘Construction Grammar’,” they still suggest— justifiably, Ibelieve— that their notation’s “ability to compactly represent large numbers of construction types” (p.110) could be asource of inspiration that construction grammarians might benefit from on their quest for an optimal system of notation as the buttress of arobust syntactic theory. Second, although not focusing on English in their contribution, the authors provide alink1 to the assembly of valency templates for English as well. Third, the applications of the notation system include the multilingual valency database MultiVal,2 certainly auseful resource. Denis Creissels and Céline Mounole’s Non-canonical valency patterns in Basque, variation and evolution provides an intriguing insight into the workings of one of the few non-Indo-European languages of Europe. Relying on data from Old Basque, Standard Basque, and present-day dialects, the authors document the expansion of ergative encoding, i.e. the expansion of coding frames lacking an absolutive slot, in the history of Basque. They attempt to provide an explanation for the rise of originally non-canonical coding frames, arguing that “the situation observed in present-day Basque implies achange in the constraints underlying the organization of the valency properties of Basque verbs or in their relative ranking” (p.151). Since the diachronic perspective on valency remains to be rather an underresearched area, this paper might be invaluable for anyone interested in the field. 1 <https://typecraft.org/tc2wiki/Valence_Profile_English> 2 <https://typecraft.org/tc2wiki/Multilingual_Verb_Valence_Lexicon> JAKUB SLáMA 243 The second part of the monograph, comprised of another four studies, covers the topic of valency rearranging alternations. Andrej L.Malchukov’s Exploring the domain of ditransitive constructions forms alink between the present volume and the Leipzig Valency Project mentioned above. Malchukov focuses on lexical variation of ditransitive constructions, illustrating that this variation is not random as there are preferences of certain verb classes for certain syntactic alignments. These preferences are shown to be valid cross-linguistically, with the author using examples from awhole gamut of languages. In the conclusion to his paper, the author points out that “more groundwork needs to be done on semantic analyses of individual verbs and constructions” (p.214), highlights the need for verb lexicons of individual languages, and concludes by calling for more progress “on both the theoretical and the empirical side to establish universals in the domain of verb classes and valency alternations” (ibid.). Chiara Melloni and Francesca Masini’s Cognate constructions in Italian and beyond identifies three types of cognate constructions (CCs) in Italian, mainly focusing on the Cognate Object Construction (COC), well-known from English, as in Sarah smiled acharming smile. The authors analyze the kinds of verbs and nouns that appear in Italian CCs, using two web corpora. Contrary to what has previously been claimed, their analysis shows that COCs take semantically heterogenous verbs as well as object nominals with divergent semantic properties. Furthermore, the authors pinpoint several lexico-semantic (aspectual properties, event schemas, noun types) and pragmatic (redundancy, and perhaps register) constraints on the use of the COC. Comparing Italian with other Romance languages, the authors identify language specific features of the use of CCs; after taking Russian and Hebrew data into account, they also suggest that the Italian CC seems to be favored in certain stylistic contexts, and that this might be alanguage specific tendency. Michela Cennamo’s Object omission and the semantics of predicates in Italian in acomparative perspective investigates the semantics of object omission with divalent verbs in Italian, paying attention to several variables and contextual features. Variability in the omissibility of the object in Italian is shown to reflect both the event structure template of the verb and semantic elements lexicalized in the verb, interacting with notions such as animacy and referentiality. While the constraints on indefinite objects (as in Anna scriveva e Ugo studiava ‘Anna was writing and Ugo was studying’) in Italian and English seem to be similar (e.g. in both languages indefinite null object are typically non-animate), anaphoric null objects appear to display ahigher degree of language specific variability, which should be further investigated. In her contribution titled On animacy restrictions for the null object in Brazilian Portuguese, Sonia Cyrino addresses some of the questions raised by the previous article, discussing anaphoric null objects in Brazilian Portuguese, as compared to American Spanish and other languages. It is argued that Brazilian Portuguese null objects are characterized by acluster of seemingly unrelated properties (including non-animate antecedents and aban on matrix subjects as antecedents). As the analysis is couched within the framework of generative syntax, the author considers null objects to be instances of DP ellipsis, i.e. inaudible DPs that have identical antecedents, which enables her to assume that “there is aparallelism requirement to be 244 LINGUISTICA PRAGENSIA 2/2018 met, and the null object and its antecedent must be in paralleled structures” (p.291). The parallelism requirement is claimed to be so strong as to be able to override the non-animacy requirement, which is supported by grammatical (albeit peripheral) examples such as Pedro beijou aMariai depois que oPaulo abraçou _i ‘Pedro kissed Maria after Paulo hugged (her).’ Nevertheless, the author concludes that the very restricted possibility of animate antecedents remains an open issue. The last part of the volume, which includes five contributions, focuses on valency changing strategies and voice alternations. Leonid Kulikov and Nikolaos Lavidas’s Between Passive and Middle focuses on Greek and Vedic verbs that can appear with two non-active morphologies. The crucial part of the contribution presents the results of adiachronic corpus study, comparing Homeric Greek, Classical Greek (Plato), and Koine Greek (New Testament), which shows “adecrease of the presence of the verbal class that can appear with two non-active voice morphologies in Greek” (p.322). This decrease might be attributable to the ongoing loss of the middle voice morphology but also to the rise of labile3 verbs that have active voice morphology for causative as well as anticausative4 uses. In Valency alternations between inflection and derivation, Livio Gaeta contrasts Italian and German, focusing on various valency changes and alternations in an effort to organize them along the Inflection-Derivation continuum. Valency alternations are usually treated either as alexical phenomenon or as acontext-conditioned variation (i.e. as apurely syntactic phenomenon); Gaeta argues for the former, which is compatible with the view of valency alternations as atype of conversion, i.e. aprocess of lexeme formation. The author thus views, for instance, bake— averb allowing for the dative alternation— as polysemous, with different senses being activated in Ibaked acake for her and Ibaked her acake. While it needs to be emphasized that the author does not necessarily plead for apurely lexicalist view of alternations, Istill believe that her discussion might have benefited if arguments such as those put forward by Goldberg (2006: Chapter 2) were taken into account. The only argument against (neo-)constructionist approaches (which are discarded without much consideration) mentioned by Gaeta is the relevance of the argument structure for derivational morphology, as exemplified by the Italian -bile, German -bar, and English -able, “which normally and productively select only transitive verbs” (p.368). However, Ibelieve that this is not really avalid argument against (neo-)constructionist approaches, since these can account for the restrictions on word-formation by the above-mentioned suffixes by invoking semantic constraints, i.e. without actually making reference to the argument structure per se (cf. the very introduction of Goldberg 1995). Ismael Iván Teomiro García’s Pronominal verbs across European languages proposes aclassification of pronominal verbs, i.e. verbs “whose syntactic realization of their argument structure requires or allows the insertion of aparticle that lacks 3 Lability means having the “same morphology for the transitive and intransitive use” (p.298). 4 Anticausative verbs are verbs that “semantically denote achange of state and enter the causative-anticausative alternation” (p.381) exemplified by John opened the door (the causative variant) and The door opened (the anticausative variant). JAKUB SLáMA 245 […] interpretation as regards the participant structure of the event” (p. 375), as in the Spanish Juan se asustó ‘Juan got scared’ or the Czech Sklo se rozbilo ‘The glass broke.’5 The proposed classification is based primarily on Spanish, but the author refers to awide variety of European languages as well, concluding with examples of questions that still need to be answered. Since there have been extensive debates over the status of the Czech word se (be it apronoun or aparticle), it could be noted that Czech linguistics might provide very good suggestions on how some of these questions might be answered (cf. Pergler 2016, for instance). The topic of reflexivity is also relevant for the following contribution, Adina Dragomirescu and Alexandru Nicolae’s Semantic constraints on the reflexive/non-reflexive alternation of Romanian unaccusatives, which discusses semantic effects associated with the alternation of aset of 38 Romanian unaccusative verbs. Even though the authors conclude that “aunitary semantic analysis of the alternation is not available for Romanian” (p.407), they still manage to delineate several subclasses of these verbs. The availability of the alternation in each of these subclasses seems to be motivated by different factors, which might include semantic differences between the reflexive and the non-reflexive variant as well as the register and stylistic specializations. By comparing Romanian with other Romance languages, the authors manifest that these factors might be both language-specific and verb class-specific. Finally, Anna Malicka-Kleparska’s Circumfixed causatives in Polish against apanorama of active and non-active voice morphology endeavors to account for the peculiar behavior of Polish circumfixed causatives, i.e. asubclass of causatives without corresponding synthetic anticausatives. For instance, while neither *ranieć nor *poranieć are attested Polish word forms, Kule mogłyby poranić konie ‘The bullets could injure the horses’ (with the circumfix underlined) is agrammatical sentence; analogous examples can be found for instance in Russian as well as in Old Church Slavonic. The author, drawing on her theoretical background of the root-based Generative morphosyntax, argues that “the limitations on the distribution of circumfixed causatives and anticausatives […] result from the particular distribution of Active voice and Non-active voice heads” (p.464). To conclude, the studies in the volume come from various functionalist and formalist theoretical backgrounds, and they explore anumber of interrelated phenomena, using data from arelatively large number of languages. While some of the papers contribute to ongoing discussions of certain issues, others manage to explore relatively underresearched topics, including afew notable diachronic studies. My overview of the individual contributions is blatantly reductive, and so Iwould like to point out explicitly that the monograph does not include asingle contribution that would not turn out quite thought-provoking and inspiring, or that would fail to prove that the cross-linguistic study of (verbal) valency is afruitful enterprise worth pursuing. The book provides an invaluable resource for researchers and students of not only general linguistics. 5 However, another Czech example provided, Jan se překvapil ‘John became surprised,’ hardly fits the definition as se seems to be areflexive pronouns coreferential with Jan; ‘John surprised himself’ would be the correct translation, Isuspect. 246 LINGUISTICA PRAGENSIA 2/2018 REFERENCES Goldberg, A.E. (1995) Constructions. AConstruction Grammar Approach to Argument Structure. Chicago and London: The University of Chicago Press. Goldberg, A.E. (2006) Constructions at Work. The Nature of Generalization in Language. New York: Oxford University Press. Hellan, L. and M.E.K.Dakubu (2010) Identifying Verb Constructions Cross-linguistically [SLAVOB Series]. University of Ghana. Retrieved on 28 December 2017 from <http://typecraft.org/w/ images/d/db/1_Introlabels_SLAVOB-final.pdf>. Malchukov, A. and B.Comrie (eds) (2015) Valency Classes in the World’s Languages. Vol.1: Introducing the Framework, and Case Studies from Africa and Eurasia. Berlin; Boston: DeGruyter Mouton. Pergler, J. (2016) Kotázce tzv.akuzativního se včeštině: pohled (nejen) diachronní. Slovo aslovesnost 77, 102–122. Tesnière, L. (1959) Éléments de syntaxe structurale. Paris: Klincksieck. Jakub Sláma Czech Language Institute of the Czech Academy of Sciences Letenská 4, 118 51 Praha 1, Czech Republic [email protected]