Grammaticalization and English Complex Prepositions

A Corpus-based Study

Sebastian Hoffmann

• Publisher: Routledge
• Year: 2007
• ISBN:
  9781134237265
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: English Language
• Index: Languages & Linguistics

What is a grammatical unit? How does grammatical structure evolve? How can we best investigate the mental representation of grammar? What is the connection between language use and language structure? This book aims to help answer such questions by presenting a detailed analysis of English complex prepositions (e.g. in spite of or with respect to) on the basis of large amounts of authentic language data dating from the Middle Ages until today.

1
Introduction

The aim of this study is to investigate the grammaticalization and use of English complex prepositions on the basis of large amounts of authentic language. I will explore corpus data spanning the time from the Middle Ages to the twentieth century and discuss the theoretical justification for classifying these constructions as a grammatical category.

Three typical examples are shown in (1) to (3):

(1) In view of what occurred later, one may believe that the young man was intoxicated. (BNC: CE9:880)

(2) In spite of the inevitable tension, it will be a relief to jump out of the aircraft. (BNC: A77:24)

(3) These hazards may lead to complications with regard to insurance. (BNC: APV: 386)

In all three cases, the italicized elements (in view of, in spite of and with regard to) are generally considered to function as heads of prepositional phrases. In other words, although they consist of three lexical items, they are still treated as single units of grammar. It will be one of the primary purposes of my study to test the validity of this claim.

English complex prepositions have received relatively little scholarly attention. There is no book-length investigation of these constructions available to date and even the number of short studies on this topic is relatively limited (cf. Quirk and Mulholland 1964; Seppänen et al. 1994; Schwenter and Traugott 1995; Akimoto 1996, 1999).¹ This lack of scholarly interest may seem all the more surprising when it is considered that Klégr’s recently published dictionary of English complex prepositions lists a total of 1,084 different preposition–noun–preposition constructions and their equivalents in Czech (Klégr 2002).² The present investigation is intended to help fill this gap by providing a detailed description of the historical development and present-day use of English complex prepositions. As I will demonstrate, such an undertaking can contribute to a better understanding of the establishment of grammatical structures in general.

In this study, I will make extensive use of data from both diachronic and synchronic corpora. The overall approach of my investigation is thus resolutely descriptive and empirical. An empirically descriptive analysis of historical data will make it possible to trace the development of complex prepositions over the last 750 years. With reference to Present-day English, the use of the 100-million word British National Corpus will allow me to investigate different aspects of current usage as well as the distributional characteristics of complex prepositions in greater detail. In contrast to a purely introspection-based approach, which relies on the intuitions of a single scholar, a corpus linguistic methodology offers access to the language output of a cross-section of speakers and can therefore reveal meaningful patterns of usage which would otherwise elude the attention of one conscious mind. I intend to show that such a corpus-linguistic description of language use can also therefore offer important insights into the structure of language as a communicative system.

The fundamental underlying assumption in such an approach is that language structure is influenced by language use. Rather than seeing language as a closed, self-contained system, language is interpreted as a dynamic entity which is constantly shaped by its users. This study will draw heavily on the concepts formulated in grammaticalization theory, which offers a useful framework for an understanding of the development of grammatical structures. In contrast to the approach taken by generative linguists, grammaticalization theory postulates a direct connection between context of use and the gradual establishment of grammatical structure. In doing so, grammaticalization theory goes beyond a mere description of language change in that it offers a (usage-based) explanation for the structures observed and their changes over time. Given this emphasis on language use, it may seem surprising that few studies of grammaticalization have so far drawn extensively on large amounts of corpus data.³ The present study of the grammaticalization of complex prepositions, however, will be based on an expansive database of authentic language use. As I will show, such an approach offers valuable insights not only into the structure under investigation but also into the nature of grammaticalization.

A further important premise underlying this study is that language use can be interpreted as a reflection of the cognitive processes at work during the production of language. Thus, if complex prepositions do indeed have the status of units of grammar, they will be stored in and retrieved from memory as whole entities rather than being assembled on the basis of the compositional rules of syntax. As I will demonstrate, the analysis of corpus data offers important insights for an investigation of the cognitive representation of grammatical structure. In spoken language, production problems such as pauses and other types of hesitation markers can be seen as manifestations of the cognitive limitations encountered during the online production of speech. On a more general level, frequency of use will be interpreted as an important indication of the mental representation of grammatical structures. If language users have at their disposal several grammatically distinct options for expressing the same concept, the frequent selection of one particular variant will be interpreted as an unconscious expression of a cognitively motivated preference. Thus, while generative grammarians are concerned with what is or is not possible within the system of grammar, I will take a predominantly usage-based approach and assign a greater level of explanatory power to the description of more or less likely choices. At the same time, however, my study will also caution against a too uncritical approach towards frequency of use. In this context, I will focus on aspects of corpus-linguistic methodology and reflect on its strengths as well as its limitations.

The current work does not propose to be comprehensive in its diachronic and synchronic study of complex prepositions, but rather to examine a particular set of prepositions in detail. This study will analyse preposition–noun–preposition constructions (PNP-constructions) such as those shown in (1) to (3).⁴ Two-word sequences such as instead of and according to, which also function as complex prepositions, will not be discussed here because there is a considerable degree of variety in two-word complex prepositions. The first item is typically an adverb (e.g. upwards of), an adjective (e.g. exclusive of), or a conjunction (e.g. as for) (cf. Quirk et al. 1985:669). The restriction to three-word sequences was adopted to minimize the impact of such formal differences on the observed findings. This will consequently allow me to take a unified approach to the constructions under consideration. For the same reason, I also excluded three-word complex prepositions which do not contain a nominal middle element (e.g. as far as, as opposed to) as well as complex prepositions in which the nominal element is preceded by a determiner (e.g. in the light of, in the face of). The bulk of this study will be concerned with an analysis of the 30 most frequent PNP-constructions in Present-day English as represented by the British National Corpus.

Chapter overview

In Chapter 2, I present a detailed description of the database that provides the empirical foundation upon which this investigation is based. A critical evaluation of the corpora used is of utmost importance since the value of corpus-linguistic research greatly depends on the degree to which the retrieved findings can be considered representative of general language use. This is particularly relevant in connection with the quotations database of the Oxford English Dictionary (OED), which constitutes a major source of data for my diachronic analysis of complex prepositions. Although the OED quotations do not fully meet certain standard requirements of language corpora (e.g. with respect to the selection criteria and its data format), my linguistic appraisal of this database will reveal that it is clearly a very valuable source of data for the study of language change. The Gutenberg Corpus, which is a collection of fiction and non-fiction texts dating from approximately 1650 to 1900 and the British National Corpus (BNC) are also described in some detail. The chapter concludes with a brief section outlining a number of methodological considerations which are relevant in the context of automated data retrieval.

Chapter 3 is concerned with the status of complex prepositions as single units of grammar. I first give an overview of what nineteenth- and twentieth-century grammars have had to say on these constructions. This is followed by a presentation of the critical view taken by Seppänen et al. (1994), who claim that the unit-like status of complex prepositions cannot be upheld when the constructions under consideration are submitted to standard constituency tests. The authors argue that complex prepositions constitute a semantically motivated category which is not helpful as part of a structural description of English grammar. In the remaining part of Chapter 3, this strong assertion is tested against actual language use found in the BNC. I will show that a usage-based and quantitative perspective on the system of language offers important insights into the nature of PNP-constructions which clearly suggest that the class of complex prepositions constitutes a meaningful element of the grammatical system.

In Chapter 4, the focus turns to grammaticalization theory as a useful framework for the description of grammatical structures. Given its ability to accommodate the gradient nature of linguistic categories, grammaticalization theory offers a more suitable basis for an evaluation of the grammatical status of complex prepositions than the categorical views held by Seppänen et al. (1994). I introduce the major concepts of this functionalist approach to the grammatical system of language by way of tracing the development of the complex preposition in view of over the last 300 years.

In Chapter 5, I offer a comprehensive overview of the diachronic development of the whole set of 30 PNP-constructions. In the course of this undertaking, I investigate whether or not the concepts introduced in the previous chapter can be meaningfully applied to the class of complex prepositions in general. I furthermore attempt to determine whether complex prepositions exhibit common features in their development towards unit-like constructions, despite the fact that they become part of the grammatical system of English at very different periods of time. I also explore whether my data supports the claim that grammaticalization is by definition a gradual process. As I will show, a considerable number of PNP-constructions in fact emerge with hardly any trace of a gradual development. As a consequence, I suggest that the grammaticalization of constructions may occur by analogy; i.e. their establishment may be greatly facilitated by their formal parallelism to previously grammaticalized items.

Chapter 6 is devoted to a description of complex prepositions in Present-day English. In a first step, I investigate the distributional characteristics of the whole set of 30 PNP-constructions over the metatextual categories annotated in the British National Corpus. This is followed by a more detailed description of the use of four individual complex prepositions. As part of the discussion of these findings, I also evaluate some of the methodological issues involved in interpreting descriptive statistics. In a final step, I focus on the concept of concession and its various formal manifestations. The use of the complex preposition in spite of is compared to its simple counterparts despite and notwithstanding.

Chapter 7 focuses on in terms of, the most frequent complex preposition in the British National Corpus. After a brief overview of its distributional characteristics in Present-day English, the bulk of the chapter is concerned with the use of this particular PNP-construction in spoken interaction. I present data which strongly suggests that in terms of has reached an advanced level of grammaticalization: it has acquired the pragmatic features of a discourse marker.

In Chapter 8, I turn to the other end of the frequency scale. While the bulk of the present investigation is concerned with a description of the 30 most frequent complex prepositions in Present-day English, this chapter concentrates on PNP-constructions with an overall frequency of between five and 100 occurrences in the 100 million words of the BNC. The main question in this context is whether or not such rare constructions can in fact be considered grammaticalized units of language. Low-frequency phenomena are normally granted only a minor role in an approach which sees language use as the instrumental factor in the shaping of language structure. After offering an overview of previous accounts of the connection between grammaticalization and frequency, I go on to suggest that the cognitive representation of grammatical structures may be influenced by factors other than frequency of occurrence alone. The final section of the chapter is devoted to the question whether or not quantitative data can be employed at all to evaluate the grammatical status of such low-frequency complex prepositions.

Chapter 9 provides a summary of the investigation and highlights my most important findings and theoretical conclusions. It also provides suggestions for future research.

2
Data collection and research methodology Introduction

The aim of this stu...

Table of contents

• Routledge advances in corpus linguistics
• Contents
• Figures
• Tables
• Acknowledgements
• Abbreviations
• 1 Introduction
• 2 Data collection and research methodology Introduction
• 3 Complex prepositions
• 4 Grammaticalization and complex prepositions
• 5 Complex prepositions
• 6 Complex prepositions in Present-day English
• 7 In terms of
• 8 Are low-frequency complex prepositions grammaticalized?
• 9 Conclusion
• Appendix I The texts of the Gutenberg Corpus
• Appendix II David Lee’s genre classification scheme
• Notes
• Bibliography
• Author index
• Subject index