LINGUIST List 28.670

Fri Feb 03 2017

Review: English; Applied Ling; Ling Theories; Socioling; Syntax: Hilpert, Östman (2016)

Editor for this issue: Clare Harshey <clarelinguistlist.org>


Date: 05-Sep-2016
From: Víctor Valdivia <vvaldiviagwu.edu>
Subject: Constructions across Grammars
E-mail this message to a friend

Discuss this message


Book announced at http://linguistlist.org/issues/27/27-1689.html

EDITOR: Martin Hilpert
EDITOR: Jan-Ola Östman
TITLE: Constructions across Grammars
SERIES TITLE: Benjamins Current Topics 82
PUBLISHER: John Benjamins
YEAR: 2016

REVIEWER: Víctor Valdivia, George Washington University

Reviews Editor: Helen Aristar-Dry

Summary

Construction Grammar represents one of the most holistic and innovative linguistic models in recent decades due to its conception of constructions as form-meaning pairings, whole units of different sorts that include different type of information, both linguistic and extra linguistic. Unfortunately, most research using Construction Grammar has focused on phenomena from single languages, mainly English. “Constructions across Grammars”, edited by Martin Hilpert and Jan-Ola Östman, moves aways from such a tendency, and thus contributes to the theory of construction grammar across languages, by presenting six articles that focus on languages other than English, but also on phenomena from different situations of language contact and on multilingual contexts. A summary of said collaboration, in the order in which they appear, is included with its evaluation in what follows.

Evaluation

In “On the borrowability of subject pronoun constructions in Turkish-Dutch contact”, the first article of the book, Seza Doğruöz, compares the occurrence of subject pronouns in Turkish spoken in the Netherlands (NL-Turkish) and Turkish spoken in Turkey (TR-Turkish). From a usage-based perspective, the author examines the role of Dutch-Turkish contact in the emergence of unconventional constructions in NL-Turkish. Even though no significant quantitative differences arise in terms of frequency between the two varieties of the language, a qualitative analysis reveals that NL-Turkish speakers often copy Dutch constructions as chunks, thus causing innovative appearances of subject pronouns. The qualitative analysis shows as well that NL-Turkish unconventional constructions can be classified into two main groups: fixed constructions and partially schematic constructions.

For the first group, Doğruöz identifies at least two constructions: ‘I don’t know’ and ‘as far as I know’. In both cases, the unconventional use of subject pronouns in NL-Turkish results from translating lexically fixed constructions from Dutch, a non pro-drop language, into Turkish, a pro-drop language. Furthermore, because the translated expressions include a subject pronoun even in contexts in which TR-Turkish speakers consider it unnecessary, the new constructions may function as fillers and evidential markers. The second group refers to syntactic schemas which normally do not include a subject pronoun, or which do not even exist in TR-Turkish. The author identifies four constructions illustrating these phenomena: subordinate constructions, left dislocations, ‘do you mean’, and the ‘yes/no’ question constructions.

Dutch influence on NL-Turkish subordinate constructions causes the occurrence of subject pronouns in cases of coreferentiality; for TR-Turkish, such unnecessary repetition of the pronoun conveys a contrastive meaning. Doğruöz hypothesizes that the inclusion of a subject pronoun in NL-Turkish copies Dutch schematic constructions; for instance, [als S V] ‘when S V’ and [dat S V] ‘that S V’. Left dislocations in NL-Turkish arise from the partial schema [NP dat/die V] ‘SNP that/those V’, which is very common in Dutch, but does not exist in TR-Turkish. Like the Dutch construction it copies, the novel NL-Turkish one introduces the topic of the following clause, in which the subject is, precisely, the referent of the left dislocation. Finally, the NL-Turkish ‘yes/no’ question construction also deviates from the norm by including a subject pronoun in non-contrastive contexts.

The paper provides clear and convincing arguments for the development of unconventional constructions in NL-Turkish, as well as its importance for the theory of construction grammar across languages. Furthermore, consulting a panel of TR-Turkish speakers to confirm the conventionality or unconventionality of the constructions minimizes the subjectivity of a qualitative analysis. As well, the section on Turks and Turkish in the Netherlands provides the reader with a clear overview of what factors interplay, and how, in the maintenance of NL-Turkish. Unfortunately, as mentioned by Doğruöz herself, results and conclusions are based on a low number of tokens, which makes difficult to anticipate whether the unconventional constructions discussed in the paper represent the beginning of a syntactic and pragmatic change or whether they illustrate a merely sporadic phenomenon.

In “On the universality of frames: Evidence from English-to-Japanese translation”, Yoko Hasegawa, Russell Lee-Goldman and Charles Fillmore use the concept of frame and the annotation system as developed in the Berkeley FrameNet project to analyze a series of English-to-Japanese translations of scientific passages. In addition to testing the applicability of the project and the universality of frames, the authors seek to answer the question of whether or not said frames can be used to assess the accuracy of translation. In the introduction to the paper, the authors describe FrameNet, define frame, explain how frames are arranged in a hierarchy, and provide examples of how to label them in the project. Although brief, the overview of the concepts is clear enough for those unfamiliar with the project. Essential as well to understand the paper’s goals and implications is the overview of typological differences in framing causal events, and how said differences often requires the use of rhetorical strategies in the translated version. This section also demonstrates how identifying and describing frames represents a more objective and systematic strategy to translation, thus providing an argument for the advantages of using this approach over other ones.

Sections 3 and 4 of the paper are based on the analysis and discussion of three source-and-translation pairs of examples; rather than working against the paper’s goal, focusing in just a few cases results in an in-depth examination of each example; for instance, the authors show how the same passage can be described using different frames, and how such description relates to the Japanese translations. Furthermore, by addressing three types of scenarios -a highly accurate one in terms of frame structure and meaning, one in which an absent frame does not seem to affect the meaning of the translation, and one in which a missing frame does affect the meaning of the translation- the authors demonstrate both the utility and limitations of using frames as a tool for assessing translations’ accuracy and, implicitly, their role in a general theory of translation. Moreover, even though the authors do not mention anything on the topic, understanding what lexical relations each particular language establishes within each frame is an essential step towards the development of accurate automatic translation applications.

In “Phonological elements and Diasystematic Construction Grammar”, Steffen Höder argues for the inclusion of phonology within the approaches of Construction Grammar (CxG), in general, and of Diasystematic Construction Grammar (DCxG), in particular. According to the author, phonological elements are typically seen as lacking motivation, which disqualifies them as constructions as understood by common definitions of the term (e.g. Goldberg 1995, 2006). On the contrary, Höder considers that features such as stress (e.g. initial vs final), presence or absence suprasegmental units (e.g. Danish stød), and tonal contours carry essential information for a construction’s form. Furthermore, Höder argues that in situations of language contact, speakers often interpret phonological elements somehow equivalent in different languages as instances of more schematic constructions; that is to say, similar language-specific features may constitute the basis for diaconstructions emerging from that particular linguistic situation. For instance, although Standard German /haʊs/ and Low German /hus/ differ in their specific realizations, they both can be seen as instances of the schema /h_s/. As convincing as the general idea of diaconstructions are, the fact that the author just compares items from Standard German and Low German may seem problematic to some scholars, particularly those studying phenomena involving contact of languages with different phonological elements.

In “Clause combining across grammars: A contrastive analysis of L1 and L2 construal of discourse organization”, Bracha Nir compares and analyzes usage of bi-clausal constructions in written personal narratives by three groups of high school and university students: native Hebrew speakers writing in Hebrew (H1H), native English speakers writing in English (E1E) and native Hebrew speakers writing in English (H1E). Following Berman’s and Slobin’s (1994) notion of syntactic packaging, Nir seeks to demonstrate that a) rhetorical constraints related to Hebrew constructions impact the usage of equivalent constructions when writing in a second language, namely English; and b) that said impact reflects the understanding, or lack of, of pragmatic constraints. Results show that although H1E speakers successfully avoid using prevalent H1H constructions when writing in English, such as linear adverbial modification, they do have problems following other English constraints such as those related to non-finite clause combinations. Findings of this sort not only illustrate how complex issues of second language acquisition are, particularly when dealing with macro-level organization of events, but also provide evidence regarding the reason that morpho-syntactic phenomena cannot be completely understood without taking into account discourse features. Even though the author does not mention what the research implies for the field of second language acquisition, it is clear that understanding how constraints from L1 affect L2 will help the development of teaching and assessments tools, particularly in advanced stages of the acquisition process in which students often reach a plateau due to the lack of strategies addressing how cultural and pragmatic aspects interplay with grammatical structures.

In “Constructional tolerance: Cross-linguistic differences in the acceptability of non-conventional uses of constructions”, Florence Perek and Martin Hilpert test the typological distinction between “constructionally tolerant” languages and “valency-driven languages”, according to which certain languages allow speakers some freedom to use a lexical item in non-conventional syntactic constructions, while other ones are less open to such syntactic innovations. In particular, the authors hypothesize that speakers of a tolerant language, e.g. English, will judge the grammaticality of an utterance produced in a second language more favorably than speakers of a less constructionally tolerant language, e.g. French. Broadly speaking, results from a grammaticality judgment task administered to speakers of the above mentioned languages support the hypothesis; however, they also make evident that the type of language per se is not the only, nor the predominant, feature driving constructional tolerance: factors such as the particular construction being judged and the existence of a similar one in the judge’s first language also have an important effect. In terms of methodology, it is worth mentioning that the authors correctly anticipate, and act accordingly, a factor that may affect the accuracy of the results: the interplay between speakers’ level of proficiency and their judgement of an utterance’s grammaticality. Taking into account such issues when factoring the scores in which the analysis is based eliminates skewed results, and allows for the study to be replicated.

In “Constructions do not cross languages: On cross-linguistic generalizations of constructions” , the last study included in the book, Philipp Wasserscheidt contends against a frequent assumption on bilingualism according to which constructions can cross from one language into another. Based on data from typologically distinct languages, the author argues that although cross-linguistic generalization may be possible according to construction grammar, it can only occur on the semantic level; thus, when speakers code-switch or seem to calque structures from another language, they do so based not on language-specific constructions, but rather on abstractions of them. The author hypothesizes that if cross-linguistic constructions were real, bilingual speakers would be able to choose what language to use in the realization of syntactic constituents. To prove such is not the case, he first illustrates his point with an array of constructions from a variety of bilingual situations, and then he focuses on prepositional phrases from the literature on bilingualism. His analysis show that the language of the preposition and the language of the case marker are the same in most cases; when both elements correspond to different languages, the marking of the noun does not correspond to the preposition. That is to say, when uttering specific elements, speakers are not transferring features from one language to another. Generally speaking, the author presents a strong argument and choose adequate examples to support it; nevertheless, given the relevance of the claim, fifty instances from various bilingual situations may not be sufficient: more data from a particular bilingual community and from different constructions may provide a better insight into what cross-linguistic transfer is and how it works.

In conclusion, by focusing on phenomena occurring in multilingual contexts, the studies included in Constructions across Grammars not only contribute to Construction Grammar and linguistics in general; they also help to fill a gap often observed in linguistic models: how general models and approaches apply to situations of bilingualism and language contact. Certainly, scholars interested in the topics and phenomena addressed in this book are looking forward to the continuation of this endeavor started by Martin Hilpert and Jan-Ola Östman. Finally, in terms of edition, it would be useful if the contributions had followed the same criteria when presenting examples and their glosses: while some authors clearly indicate what particular element the reader needs to pay attention to (e.g. making the text bold) others do not do so. Certainly, the latter does not cause misunderstanding of the particular utterance, but readers unfamiliar with the language will appreciate the help when examining the example.


ABOUT THE REVIEWER

Professor of Spanish at the George Washington University. My research interests include functional syntax, particularly in situations of language contact.

Page Updated: 03-Feb-2017