Date: Tue, 28 Jun 2005 10:51:34 +0100 From: Nicole Dehé Subject: English Ditransitive Verbs
AUTHOR: Mukherjee, Joybrato TITLE: English Ditransitive Verbs SUBTITLE: Aspects of Theory, Description and a Usage-based Model SERIES: Language and Computers Vol. 53 PUBLISHER: Rodopi YEAR: 2005
Nicole Dehé, Department of Phonetics and Linguistics, University College London
Joybrato Mukherjee's book provides an addition to the literature on ditransitive verbs in English. Based on real language corpus data, the study offers a meticulous description of the usage of six ditransitive verbs in particular (give, tell, send, show, ask, offer). The descriptive analysis is embedded in a pluralist theory developed on the basis of previous accounts to ditransitive verbs put forward in various frameworks and it leads to the development of a usage-based model of ditransitive verbs combining a general cognitive-linguistic framework with specific corpus-linguistic considerations.
The book is divided into five well-organised chapters plus a short appendix providing details on the design of the corpus used in the study.
EVALUATION BY CHAPTER
Chapter 1 is intended to give an overview and critical review of different existing frameworks in various traditions used for the description and analysis of ditransitivity and their different perspectives, among them descriptive grammar, generative grammar, valency theory, functional grammar and semantico-syntactical approaches, corpus-based grammar including some previous corpus- based research into ditransitive verbs, corpus-driven lexicogrammar, construction grammar and cognitive sciences. Based on this overview, a pluralist theory of ditransitivity is developed from those aspects of the frameworks discussed that "can be conceptually integrated with each other" (p. 63). Moreover, a working definition of ditransitive verbs is provided. According to this definition, ditransitive verbs are trivalent verbs which require a subject, a direct object and an indirect object. The pattern where both objects appear as noun phrases (as in 'Peter gave Mary a book') is referred to as the "basic" or "explicit" ditransitive syntax. In what follows, all (and only) verbs that meet this definition, i.e. which are attested in the basic pattern at least once in the data set, are considered ditransitives in the basic as well as all other forms of complementation.
At this point, a remark concerning the sections on previous literature seems in order. Despite providing a useful overview in general, it is striking that while all other approaches are dealt with in a neutral and objective way, the section on generative grammar reads more like a misplaced attack on the generative framework in general than a critical and fair discussion of approaches to ditransitives suggested within that framework. Traces of this attack can be found throughout the book. While it might be true that it is quite a challenge to reconcile the generative framework (used here synonymously with "Minimalist Program") with a usage-based model as suggested in the present study, a more objective discussion would have been more fruitful. Some inaccuracies in the outline of the theory contribute to this impression (e.g. on p. 17 it is claimed that "the object [...] is a projection of the verb", and on pp. 20/21 the same construction, V NP NP, is represented by two different syntactic structures for no apparent reason).
Chapter 2 is intended to outline the methodology used in the present study which is described as a "corpus-to-cognition approach". Section 2.1 sketches the corpus-based approach to ditransitivity as one where previous research results and experience- and intuition-based hypothetical models are applied to actual performance data, leading to models (of ditransitivity) in language use and language cognition, and in turn to the verification or falsification of the initial hypotheses (cf. Fig. 2-1, p. 71). The author distances himself and his work from the tradition of "corpus-driven" linguistics in which the starting point for any linguistic theory is the corpus itself and he rejects the corresponding framework as "unrealistic" and "implausible" (p. 72). In Section 2.2, the relevant corpora and corresponding software tools are being presented. The main corpus used, introduced in some detail together with a review of general advantages and downsides for the present study, is the International Corpus of English - The British Component (ICE-GB) along with the ICE Corpus Utility Program (ICECUP; cf. also the corpus webpage at usage/ice-gb/index.htm> and Nelson et al. 2002). From this corpus, a list of all verbs that meet the working definition of ditransitivity as suggested in Chapter 1 was automatically derived. This list includes 70 verbs of varying frequency, adding up to 1741 occurrences of verbs parsed as ditransitive in ICE-GB. Particularly frequent are the verbs ASK (91 occurrences), GIVE (562), OFFER (54), SEND (79), SHOW (84) and TELL (491). The corpus was then manually searched for these ditransitive verbs in other complementation patterns.
Based on the overall frequency of each ditransitive verb in the corpus and on the frequency with which each ditransitive verb occurs in the basic ditransitive pattern (both obtainable from ICECUP, Table 2-5 on p. 84), the author distinguishes three groups of ditransitive verbs. Firstly, "typical ditransitive verbs" are frequent in general and in explicit ditransitive syntax. These criteria are met by GIVE and TELL. The second group is that of "habitual ditransitive verbs" which are fairly frequent in general but not in explicit ditransitive syntax, specifically ASK, SEND, SHOW and OFFER. The third group is made up of "peripheral ditransitive verbs" which are rare in general and/or in explicit ditransitive syntax. Along with ICE-GB, the British National Corpus (BNC, 2nd version), was used as "ancillary corpus" for the description of ditransitive verbs that were peripheral in ICE-GB.
Section 2.3 deals with the corpus-based description of language use. According to the present study, a model of language use is a "supraindividual abstraction" of what is frequent and/or normal in a given speech community. The author underlines that this does not equal performance, since performance data (of which a corpus is representative) includes all kinds of errors, unacceptable language use, etc. The remainder of the chapter emphasizes the main chores of the usage-based model to be developed on the basis of the corpus data: (i) to account for the fundamental importance of the actual frequencies of linguistic forms; (ii) to explain linguistic routines and patterns; (iii) to take into account the role of functional and context- dependent principles and factors; (iv) to keep in mind the interdependence of lexical and grammatical choices.
In Chapter 3, the author provides a detailed corpus-based description as well as quantitative and qualitative analysis of individual ditransitive verbs in actual language use. He addresses the three distinct groups of verbs identified in Chapter 2 in turn (Section 3.1: typical ditransitive verbs: GIVE and TELL; Section 3.2: habitual ditransitive verbs: SEND, SHOW, ASK, OFFER; Section 3.3: peripheral ditransitive verbs). These verbs occur in various syntactic pattern meticulously described in this chapter. Among these patterns are the basic ditransitive pattern ("Peter gave Mary a book"; here: pattern I), the prepositional pattern ("Peter gave the book to Mary"; here: pattern II), a mono-transitive pattern realizing the direct object ("Peter gave an example"; here: pattern III), an intransitive pattern ("Peter always gives"; here: pattern IV), a mono-transitive pattern realizing the indirect object ("Peter told him"; here: pattern V), and variations of these patterns derived by syntactic operations involving the direct and/or indirect object, such as passivisation, fronting in relative clauses and embedded questions, among others. For each verb considered here, a default pattern is identified mostly on quantitative grounds, but also on the basis of underlying semantics (in the case of ASK) and discourse patterns (in the case of TELL). The default patterns vary across items, except that the default for the habitual verbs SHOW, SEND and OFFER is pattern III.
Special emphasis is placed on the identification of factors causing the speaker to choose a specific pattern over other possible patterns. Among these factors are the explicitation and linear arrangement of semantic roles, textual factors, lexical constraints, heaviness of constituents, and pragmatic factors such as end-focus.
In some cases it seems as if the author has to bend his own rules, which occasionally leads to contradictory assumptions. To give but one example: as argued in the present study (as well as much previous work) one of the factors contributing to the choice of construction is (final) focus. Now consider the examples below. - (ex. 279, p. 159): In the other universities abroad the students are shown to have more life within them. (ICE-GB:W1A-018 #74) - (ex. 407, p. 196): Since the Theatre is a Department of University College London we only offer this casual work to students.
For the example in (279), Mukherjee (p. 159) argues that the indirect object "the students" is passivised because "the students are evoked by universities" in the immediately preceding context. However, for the example in (407), he argues (p. 196f.) that "students" represents new information and is therefore structurally realised in end-position, triggering the prepositional pattern. In my view, it remains unclear from the two examples given here, why "universities" in (279), but not "University College" in (407) evokes "students". Moreover, in many cases the argumentations as to why one pattern has been chosen over another leaves the reader with the impression that a different pattern would have done the same job. This may sound petty but is crucial since the usage-based model developed in Chapter 4 is based partly on these selection principles.
Despite these drawbacks, and even though some of the factors do not seem completely convincing to me, these sections give a thorough and to my knowledge unique overview of the actual usage of selected ditransitive verbs. While both the idea that the choice of a specific pattern is principle-guided rather than optional and the identification of guiding factors are not unprecedented in the literature, the wealth of data and the kind of frequency-based analysis offered on the basis of these data are new.
The section on peripheral ditransitives starts off from the question why the relevant verbs can be used ditransitively in the first place. Here, the author argues along the lines of grammatical institutionalization (a process in the course of which verbs whose underlying semantic component includes a potentially ditransitive meaning are licensed to become possible ditransitive verbs structurally) and conventionalization (a process by which possible ditransitive verbs, i.e. the output of grammatical institutionalization, are turned into probable ditransitive forms). More precisely, a potential ditransitive verb is not attested in the basic ditransitive pattern, but the verbal meaning includes the ditransitive situation schema either explicitly or metaphorically. Peripheral verbs are grammatically institutionalized verbs, i.e. verbs with potential ditransitive meaning are infrequently attested in the basic ditransitive structure. Examples are SUPPLY (as in 'Supply us some more drinks') and PROFIT. A verb that is used increasingly in the ditransitive pattern then turns into a habitual or typical (= conventionalized) ditransitive verb. The earlier classification between peripheral, habitual and typical ditransitive verbs is thus translated into different stages in the process of grammaticalisation, without making a distinction between the latter two verb classes along these lines.
Chapter 4 proposes a network-like "model of speakers' linguistic knowledge about ditransitive verbs" based on the corpus data and analysis presented in Chapter 3, and combining "the general cognitive- linguistic framework [...] with specific corpus-linguistic considerations" (p. 221). This chapter would have highly benefited from a more straight-to-the-point-like discussion. The first two sections develop some general aspects of the kind of model employed here, focusing on the importance of a large corpus of authentic data, as well as the integration of lexical and syntactic aspects. The reader is also introduced to one of the limitations of the methodology employed in this study, which is that due to corpus limitations it can only account for the core but not the periphery of a given phenomenon. In the course of the core/periphery discussion in Section 4.2, the three classes of ditransitives defined above are translated into "three zones of prototypicality within the category of ditransitivity", where peripheral ditransitives correspond to the least prototypical members, while typical ditransitives correspond to the most prototypical members of their class. In what follows, usage-based models are developed for the most prototypical members only. At the heart of the discussion are usage-based models of the "cognitive representation of the lexicogrammar" of the two typical ditransitive verbs GIVE and TELL. These models are highly specific, taking into account the frequency of every pattern attested for these verbs as well as the selection principles as identified for each of these patterns. Being truly usage- based, they aim at bridging the gap between large amounts of authentic corpus data and their analysis on the one hand and a competence-related model of language cognition on the other hand.
Chapter 5 serves as a summary and conclusion.
In my view, the merit of this study, Chapters 3 and 4 in particular, lies clearly in the pattern description and quantitative analysis of a wealth of authentic data. One major point of criticism that Mukherjee puts forward with respect to other models, in particular regarding the generative framework, is that these models are often based exclusively on "invented and decontextualised" (p. 229) sentences. His answer is a model that is exclusively based on real data. Clearly and fully recognized in the present study, this model is core-centered and cannot account for any processes at the periphery. Moreover, since the model focuses on individual verbs "for practical reasons", a comprehensive approach accounting also for differences and similarities between members of the class of ditransitive verbs, concerning e.g. default patterns, predominant selection principles and the like, cannot be offered (although an attempt to illustrate the network-like character of the model is provided in Fig. 4-6, p. 248).
On a more formal note: the book contains far too many footnotes and paragraph-long quotations interrupting the flow of the text, as well as some lengthy sections summarizing previous research which could often have been replaced by references to the work in question.
Leaving the advantages and drawbacks of different theoretical frameworks as well as some aspects of textual organisation aside, the data description and quantitative analysis that Mukherjee offers in this study will be of interest to any linguist who works on ditransitive verbs in English. The discussion centering on the importance of a methodology based on authentic data is of even more general interest.
Nelson, Gerald, Sean Wallis & Bas Aarts (2002) Exploring Natural Language: Working with the British Component of the International Corpus of English. Amsterdam: John Benjamins.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Nicole Dehé is an Honorary Research Fellow at UCL, Dept of Phonetics and Linguistics. She received her PhD in 2001 from the University of Leipzig. She is the author of Particle Verbs in English: Syntax, Information Structure and Intonation, 2002, Amsterdam: Benjamins, and co-editor of a volume on particle verbs (Verb-Particle Explorations, 2002, Berlin/New York: Mouton de Gruyter; with Ray Jackendoff, Andrew McIntyre and Silke Urban).