Anda di halaman 1dari 191

The M.A.K.

Halliday Library Functional Linguistics Series

Qingshun He
Bingjun Yang

Absolute Clauses
in English from the
Systemic Functional
Perspective
A Corpus-Based Study
The M.A.K. Halliday Library Functional
Linguistics Series

Series editors
Chenguang Chang
Guowen Huang
About the Series

The M.A.K. Halliday Library Functional Linguistics Series focuses on studies


­concerning the theory and application of Systemic Functional Linguistics (SFL).
As a functional theory of language, SFL was initially developed by Professor
M.A.K. Halliday and his colleagues in London during the 1960s, and since then its
influence has spread all over the world.
Systemic Functional Linguistics distinguishes itself as a functional theory by
the emphasis placed on system in relation to structure. It has also been particu-
larly concerned with modelling language in context. The theory is especially well-
known for the work on discourse analysis, cohesion, genre and register, appraisal
and so on, which have been taken up by scholars working in other fields.
Since Halliday’s early work on Chinese and English, systemic functional lin-
guists around the world have been increasing the coverage of the description of
different languages over the decades, including French, Spanish, Portuguese,
German, Danish, Finnish, Persian, Thai, Vietnamese, Japanese, Korean, Tagalog,
Bahasa Indonesian, Gooniyandi and others.
Systemic Functional Linguistics is also characterized as an “appliable” lin-
guistics theory. It is well-known for its application in a variety of fields, includ-
ing education, translation studies, computational linguistics, multimodal studies,
healthcare, and scholars are exploring new areas of application.
The M.A.K. Halliday Library Functional Linguistics Series is an open series.
Monographs included in this series will cover studies on language and context,
functional grammar, semantic variation, discourse analysis, multimodality, register
and genre analysis, educational linguistics, etc. Manuscripts are selected, based on
quality and significance, in consultation with an editorial board which consists of
leading linguists in the SFL field.

More information about this series at http://www.springer.com/series/13311


Qingshun He · Bingjun Yang

Absolute Clauses in English


from the Systemic Functional
Perspective
A Corpus-Based Study

13
Qingshun He Bingjun Yang
Faculty of English Language and Culture School of Foreign Languages
Guangdong University of Foreign Studies Shanghai Jiao Tong University
Guangzhou Shanghai
China China

ISSN  2198-9869 ISSN  2198-9877  (electronic)


The M.A.K. Halliday Library Functional Linguistics Series
ISBN 978-3-662-46366-6 ISBN 978-3-662-46367-3  (eBook)
DOI 10.1007/978-3-662-46367-3

Library of Congress Control Number: 2015931449

Springer Heidelberg New York Dordrecht London


© Springer-Verlag Berlin Heidelberg 2015
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained
herein or for any errors or omissions that may have been made.

Printed on acid-free paper

Springer-Verlag GmbH Berlin Heidelberg is part of Springer Science+Business Media


(www.springer.com)
Acknowledgments

This research could not have been completed without the help of many col-
leagues and friends, among whom we are particularly grateful to Profs. Guowen
Huang and Chenguang Chang at Sun Yat-sen University. Professors Kaibao
Hu (Shanghai Jiao Tong University) and Binli Wen (Guangdong University of
Foreign Studies) generously supported us in many ways during the writing and
revising of the manuscript. Many thanks also to Rebecca Zhu, Yi Xu and Evelyn
Ebina J. from Springer for their help. The research is supported by the Publication
Fund of Guangdong University of Foreign Studies. This research is also supported
by the Program for New Century Excellent Talents in University from China
Ministry of Education (NCET-11-0704) and the Fundamental Research Funds for
the Central Universities (SWU1409102).

Bingjun Yang

v
Contents

1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Research Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Purpose of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Organization of the Book. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 Absolute Clauses in the Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7


2.1 What Is Absolute Clause . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Types of Absolute Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.1 Free Adjunct Construction. . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2.2 Nominative Absolute Construction . . . . . . . . . . . . . . . . . . . . 13
2.2.3 Augmented Absolute Construction . . . . . . . . . . . . . . . . . . . . 14
2.2.4 Differences and Similarities. . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 Logical Roles of Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.4 The Case of Absolute Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.5 Stylistic Effects of Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . . . . 24
2.6 Questions to Be Answered. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3 Approaching Absolute Clauses from the SFL Perspective . . . . . . . . . . 31


3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Functional Structure of Clause. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2.1 Transitivity Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2.2 Mood Structure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
3.2.3 Thematic and Information Structures. . . . . . . . . . . . . . . . . . . 35
3.3 Functional Structure of Clause Complex. . . . . . . . . . . . . . . . . . . . . . 35
3.4 Rank Status of Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3.4.1 Cline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.4.2 Rank Status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

vii
viii Contents

3.5 Formation Requirement and Identification Criteria


of Absolute Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.1 Formation Requirement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.2 Identification Criteria. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

4 Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.1 Research Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 Research Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.3 Data Collection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.4 Data Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

5 Relationships Realized by Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . 65


5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
5.2 Relation Potential of Absolute Clauses . . . . . . . . . . . . . . . . . . . . . . . 66
5.2.1 Relation System Network. . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.2.2 Relation Potential. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.3 Dependent Absolute Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3.1 Expansion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
5.3.2 Projection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.4 Embedded Absolute Clauses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.4.1 Classification of Non-finite Verbs . . . . . . . . . . . . . . . . . . . . . 84
5.4.2 Types of Embedding. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.5 Continuing Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.6 Independence of Absolute Clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.6.1 Dimensions of Independence of Absolute Clauses . . . . . . . . 98
5.6.2 Independent Tendency of Absolute Clauses. . . . . . . . . . . . . . 100
5.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

6 Absolute Clauses Distributed in Three Corpora . . . . . . . . . . . . . . . . . . 107


6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
6.2 Research Based on Brown Family Corpora. . . . . . . . . . . . . . . . . . . . 108
6.2.1 Overall Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.2.2 Functional Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
6.2.3 Personal Pronoun Subject and Its Case Choice. . . . . . . . . . . 116
6.2.4 Built-in Problems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
6.3 BNC-Based Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
6.3.1 Overall Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.3.2 Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
6.4 COHA-Based Observations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.4.1 Functional Distribution of Overall Frequency. . . . . . . . . . . . 132
6.4.2 Historical: Overall Frequency . . . . . . . . . . . . . . . . . . . . . . . . 135
Contents ix

6.4.3 Function Types. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136


6.4.4 The Case of Personal Pronoun Subject . . . . . . . . . . . . . . . . . 137
6.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

7 Discussions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.1 Function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.2 Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
7.3 Style. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7.4 Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
7.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

8 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.1 Main Findings of This Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
8.2 Limitations and Further Research . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.2.1 Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
8.2.2 Further Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

Appendix A: TreeTagger POS Tagset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

Appendix B: Tables of Corpus Data Statistics. . . . . . . . . . . . . . . . . . . . . . . 167

Appendix C: Corpus Retrieving Demonstration. . . . . . . . . . . . . . . . . . . . . 175


Abbreviations

BROWN The Brown Corpus (American English, 1961)


FROWN The Freiburg-Brown Corpus (American English, 1992)
CROWN The China-Brown Corpus (American English, 2009)
LOB The Lancaster-Oslo/Bergen Corpus (British English, 1961)
FLOB The Freiburg-Lancaster-Oslo/Bergen Corpus (British English, 1991)
CLOB The China-Lancaster-Oslo/Bergen Corpus (British English, 2009)
BNC The British National Corpus
COCA The Corpus of Contemporary American English
COHA The Corpus of Historical American English
CLAWS Constituent-Likelihood Automatic Word Tagging System
CLAWS7 The newest version of CLAWS
SFL Systemic Functional Linguistics
SPOK Spoken
FIC Fiction
MAG Popular Magazines
NEWS Newspaper
ACAD Academic
NF Non-fiction
NA Non-academic
MISC Miscellaneous

xi
Notational Conventions

1 The initiative clause


2 The continuing clause
α The dominant clause
β The dependent clause
= Elaborating
+ Extending
× Enhancing
“ Locution
‘ Idea

xiii
Figures

Fig. 2.1 Core component of predicate of absolute clauses


(Kortmann 1991: 10). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Fig. 3.1 Traditional and functional structures. . . . . . . . . . . . . . . . . . . . . . . 32
Fig. 3.2 Transitivity structure of absolute clauses. . . . . . . . . . . . . . . . . . . . 34
Fig. 3.3 Mood structure of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . 34
Fig. 3.4 Thematic and information structures of absolute clauses. . . . . . . 35
Fig. 3.5 The system network of clause complexes. . . . . . . . . . . . . . . . . . . 36
Fig. 3.6 Cline from clause to group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Fig. 4.1 Six matching corpora in Brown Family. . . . . . . . . . . . . . . . . . . . . 57
Fig. 4.2 Styles and number of texts in each corpus of Brown Family . . . . 57
Fig. 5.1 Type system network of absolute clauses in traditional grammar. . . 66
Fig. 5.2 Type system network of absolute clauses in SFL . . . . . . . . . . . . . 67
Fig. 5.3 Relation potential system of absolute clauses. . . . . . . . . . . . . . . . 68
Fig. 5.4 Extension system network. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
Fig. 5.5 Extension system network of non-finite clauses. . . . . . . . . . . . . . 72
Fig. 5.6 Extension system network of absolute clauses . . . . . . . . . . . . . . . 73
Fig. 5.7 System of projection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Fig. 5.8 Classification of non-finite verbs (I) . . . . . . . . . . . . . . . . . . . . . . . 85
Fig. 5.9 Classification of non-finite verbs (II). . . . . . . . . . . . . . . . . . . . . . . 85
Fig. 5.10 Syntactic relations realized by embedded clauses. . . . . . . . . . . . . 87
Fig. 5.11 Cline of embedding depth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Fig. 5.12 Relation system of “clause + clause” structures. . . . . . . . . . . . . . 93
Fig. 5.13 Relation system of “clause + non-finite clause” structures. . . . . . 93
Fig. 5.14 System of types of absolute construction . . . . . . . . . . . . . . . . . . . 98
Fig. 6.1 Stylistic distribution of absolute clauses (per million words). . . . 112
Fig. 6.2 Historical distribution of functions of absolute clauses. . . . . . . . . 114
Fig. 6.3 Historical distribution of functions of absolute clauses
in British English. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Fig. 6.4 Historical distribution of functions of absolute
clauses in American English. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Fig. 6.5 Stylistic distribution of functions of absolute clauses. . . . . . . . . . 116

xv
xvi Figures

Fig. 6.6 Stylistic distribution of personal pronoun subjects


of absolute clauses (per million words). . . . . . . . . . . . . . . . . . . . . 118
Fig. 6.7 Stylistic distribution of the case of personal pronoun
subjects of absolute clauses (per million words). . . . . . . . . . . . . . 121
Fig. 6.8 Comparison between stylistic distributions of absolute
clauses and absolute clauses with personal pronoun subject. . . . . 123
Fig. 6.9 Stylistic distribution of absolute clauses in BNC
(per hundred million words) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Fig. 6.10 Comparison of stylistic distribution of absolute
clauses between the Brown Family Corpora and BNC. . . . . . . . . 126
Fig. 6.11 Tendency of with constructions to form absolute
clauses of extension. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Fig. 6.12 The ratio of the absolute clauses of extension
to with constructions (equal totality). . . . . . . . . . . . . . . . . . . . . . . 129
Fig. 6.13 Tendency of enhancing clauses to form absolute
clauses of enhancement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Fig. 6.14 Ratio of absolute clauses of enhancement to subordinators
(equal totality). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Fig. 6.15 Tendency of that clauses to form absolute clauses of projection. . . 131
Fig. 6.16 Ratio of absolute clauses of projection to that clauses
(equal totality). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Fig. 6.17 Ratio of absolute clauses to linkers (equal totality). . . . . . . . . . . . 132
Fig. 6.18 Functional distributions of absolute clauses of expansion
in the three corpora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Fig. 6.19 Functional distributions of absolute clauses of projection
in the three corpora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
Fig. 6.20 Overall historical distribution of absolute clauses
in COHA (equal totality). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Fig. 6.21 Overall historical distribution of absolute clauses
in COHA (per hundred million words). . . . . . . . . . . . . . . . . . . . . 136
Fig. 6.22 Historical distributions of functions of absolute clauses
in COHA (per hundred million words). . . . . . . . . . . . . . . . . . . . . 137
Fig. 6.23 Proportions of the case distribution of absolute clauses
with personal pronoun subject in COHA. . . . . . . . . . . . . . . . . . . . 138
Fig. 6.24 Functional distribution of the case of absolute clauses. . . . . . . . . 139
Fig. 6.25 Historical distribution of the case of absolute clauses
of enhancement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
Fig. 6.26 Historical distribution of the case of absolute clauses of extension. . . 141
Fig. 6.27 Historical distribution of the case of absolute clauses of elaboration. . . 141
Fig. 7.1 Proportion of absolute clauses used independently in COHA. . . . 145
Fig. 7.2 Proportion of accusatives to case-marked personal
pronoun subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
Fig. 7.3 Historical distributions of weather permitting/being…
with and without the . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Figures xvii

Fig. 7.4 Historical distribution of absolute clauses with


and without article in COHA (per hundred million words). . . . . . 156
Fig. 7.5 Historical distribution of grammaticalized absolute
clauses of extension (per hundred million words). . . . . . . . . . . . . 156
Tables

Table 2.1 Names of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9


Table 2.2 Logical roles of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . 16
Table 3.1 Basic types of clause complex (Halliday 1994: 220). . . . . . . . . . 37
Table 4.1 Styles and number of texts in the Brown Family Corpora. . . . . . 58
Table 4.2 Number of words of the Brown Family Corpora. . . . . . . . . . . . . 59
Table 4.3 COCA and BNC compared in terms of register
balance and number of words . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Table 4.4 Registers, number of words, and historical distribution
in COHA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
Table 6.1 Formal distribution of absolute clauses of expansion . . . . . . . . . 110
Table 6.2 Formal distribution of absolute clauses of projection . . . . . . . . . 110
Table 6.3 Historical and regional distribution of absolute clauses. . . . . . . . 110
Table 6.4 Stylistic distribution of absolute clauses in the Brown
Family Corpora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Table 6.5 Regional distribution of functions of absolute clauses. . . . . . . . . 113
Table 6.6 Historical distribution of functions of absolute clauses. . . . . . . . 114
Table 6.7 Stylistic distribution of functions of absolute clauses. . . . . . . . . 116
Table 6.8 Historical distribution of personal pronoun subjects
of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
Table 6.9 Stylistic distribution of personal pronoun subjects
of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Table 6.10 Regional distribution of the case of personal pronoun
subjects of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
Table 6.11 Regional distribution of nominative and accusative
personal pronoun subject of absolute clauses . . . . . . . . . . . . . . . 119
Table 6.12 Historical distribution of the case of personal pronoun
subjects of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Table 6.13 Stylistic distribution of the case of personal pronoun
subjects of absolute clauses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Table 6.14 Stylistic distribution of the case of personal pronouns
in the Brown Family Corpora. . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

xix
xx Tables

Table 6.15 Distribution of absolute clauses in the seven styles in BNC. . . . 124
Table 6.16 Stylistic distributions of with and absolute clauses of extension. . . 127
Table 6.17 Stylistic distributions of subordinators and absolute
clauses of enhancement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Table 6.18 Stylistic distributions of complementizer that
and absolute clauses of projection. . . . . . . . . . . . . . . . . . . . . . . . 127
Table 6.19 Stylistic distributions of with and absolute clauses
of extension (per hundred million words) . . . . . . . . . . . . . . . . . . 128
Table 6.20 Stylistic distributions of subordinators and absolute
clauses of enhancement (per hundred million words). . . . . . . . . 128
Table 6.21 Stylistic distributions of complementizer that and absolute
clauses of projection (per hundred million words) . . . . . . . . . . . 128
Table 6.22 Stylistic distributions of linkers and absolute clauses. . . . . . . . . 132
Table 6.23 Historical distribution of the absolute clauses in COHA. . . . . . . 133
Table 6.24 Functional distribution of absolute clauses . . . . . . . . . . . . . . . . . 134
Table 6.25 Case distribution of personal pronoun subjects of absolute
clauses in COHA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
Table 6.26 Functional distribution of the case of absolute clauses. . . . . . . . 139
Table 6.27 Functional distribution of the case of absolute clauses over time. . . 140
Table 7.1 Absolute clauses used independently. . . . . . . . . . . . . . . . . . . . . . 144
Table 7.2 Absolute clauses used independently in COHA . . . . . . . . . . . . . 145
Table 7.3 Case-marked personal pronoun subjects. . . . . . . . . . . . . . . . . . . 148
Table 7.4 Absolute clauses of enhancement in three corpora . . . . . . . . . . . 154
Table 7.5 Most frequent absolute clauses of enhancement in COHA. . . . . 155
Chapter 1
Introduction

1.1 Research Background

Absolute nominative clause is a syntactic phenomenon, which has a long history.


It began to appear in Anglo-Saxon, known as dative absolutes, which, however,
are generally considered being originated from Latin ablative absolutes. This is
because absolute nominative clauses were frequently used in the Anglo-Saxon lit-
eral translations from Latin but less frequent in free translations and even seldom
appeared in the Anglo-Saxon works. There were few absolute nominative clauses
in early Middle English, but in Chaucer’s poems in the latter half of the fourteenth
century, there were a lot of absolute nominative clauses (Ross 1893: 252–253).
According to Ross (1893), in early Modern English, namely the sixteenth and first
half of seventeenth centuries, absolute nominative clauses began to flourish in
some works of the classical style. In the late seventeenth century, absolute nomi-
native clauses became naturalized by the Restoration and were popularly used in
various styles of works because the styles of writing were no longer lengthy and
cumbersome in the writers’ pursuit of clearness and simplicity. This is in agree-
ment with the formal characteristics of absolute nominative clauses.
Absolute nominative clauses have also been the concern of grammarians. They
mainly concentrate on the case choice, formal classification, and stylistic fea-
tures of absolute nominative clauses. Case is “a grammatical category used in the
analysis of word-classes (or their associated phrases) to identify the syntactic rela-
tionship between words in a sentence, through such contrasts as nominative, accu-
sative, etc.” (Crystal 2008: 66). Absolute nominative clauses in different languages
use different cases: locative in Sanskrit, genitive in Greek, ablative in Latin, accu-
sative in French, nominative or accusative in Italian, and dative in Norman. In
Anglo-Saxon, the normal case of absolute nominative clauses was dative, just as
ablative in Latin. From the historical and analogical perspective, absolute nomi-
native clauses in Middle and Modern English should have been the oblique case.

© Springer-Verlag Berlin Heidelberg 2015 1


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_1
2 1 Introduction

However, absolute nominative clauses in late Middle and Modern English were
nominative case. “About the middle of the fourteenth century, the nominative
began to replace the dative.” (Morris 1886[2010]: 103), while the dative “lasted
down to 1400” (Oliphant 1878: 408). In fact, Wycliffe ever used nominative pro-
nouns in his translation, indicating that the case of absolute nominative clauses
had begun to change into nominative before Wycliffe, which was probably com-
pleted in 1420s or 1430s (Poutsma 1929: 973). Modern English grammarians gen-
erally hold that the subjects of absolute nominative clauses are zero case nouns
or nominative pronouns, sometimes accusative pronouns (Curme 1931: 154;
Jespersen 1949: 45; Visser 1972: 1148), but very few in number, some of which
are obviously affected by Latin. Some other grammarians think that the accusative
form is often considered unacceptable in Modern English (Fowler 1965: 4; Stump
1985: 11). Such grammarians as Quirk et al. (1985) and Biber et al. (1999) use
“absolute clauses” to name absolute nominative clauses to avoid the case prob-
lem. To some extent, they accept the fact that the subjects of absolute nominative
clauses can be accusative.
Absolute nominative clauses began to appear in English poems in the four-
teenth century and became widely used in the seventeenth and eighteenth centu-
ries, but were gradually reduced in the nineteenth century and then confined to
formal styles (Ross 1893: 14–15). As to style, it is generally believed that abso-
lute nominative clauses “is not only uncolloquial, but is by many felt to be un-
English, and to be avoided in writing as well” (Sweet 1903: 124). “Except for
stereotyped phrases like weather permitting, the colloquial use of the Nominative
Absolute is almost restricted to it being…, there being…” (Onions 1905[2010]:
67). Grammarians nowadays seldom carry out specialized research on absolute
nominative clauses and only discuss them in grammar books as a certain gram-
matical structure. However, absolute nominative clause is a very common gram-
matical structure (Wooley 1920; Pence and Emery 1965) and absolute nominative
clauses are the most commonly used sentence modifiers (Al-Hamash and Abdulla
1979: 379) in actual English usages.
Traditional grammarians are also interested in the function types of absolute
nominative clauses. Curme (1931: 154–157) distinguishes six broad logical roles
according to whether an absolute is intuitively linked to its superordinate clause,
including the relations of time, cause, condition and exception, attendant circum-
stance, manner proper, and concession. Kruisinga (1932: 274–275) assumes the
whole range of logical roles under the notion of attendant circumstances. Jespersen
(1949: 61–64) suggests that absolutes can be seen to play four sorts of logical
roles: cause or reason, condition, time, and descriptive circumstances, admitting “it
is often difficult or even impossible to draw sharp boundaries between the several
applications.” Visser (1972) categorizes four logical roles played by absolute nomi-
native clauses: (1) attendant circumstances, (2) reason, ground, cause, or motive,
(3) time, and (4) condition. Quirk et al. (1973: 762) distinguish such three types
of logical relations in absolute clauses as cause, time, and circumstance. Although
grammarians distinguish different function types of absolute nominative clauses,
they all categorize absolute nominative clauses into adverbial clauses.
1.1  Research Background 3

The above-mentioned aspects of research on absolute clauses indicate that this


traditional syntactical structure has attracted full attention of grammarians on the
one hand, and on the other hand, it also reflects the deficiencies of descriptions
within traditional grammar. Systemic functional linguistics (SFL) is meaning-
based and regards language as a meaning system. According to SFL, form is the
realization of meaning, and studies should deal with language in use, focusing
on the realization forms of meaning at the lexico-grammatical level. Traditional
grammar and SFL both discuss the relationship between grammar and meaning
(Halliday and Matthiessen 1999: 15, 28). However, the former focuses on lan-
guage form, taking words as the starting point to interpret grammatical categories
from below, which facilitates the representation of overt categories of grammar,
and the latter focuses on language function, taking clauses as the starting point
to interpret grammatical categories from above, which facilitates the exploration
of cryptotypes. Both are syntactic, but “Formal syntax deals with how words can
combine to create larger units of form and eventually sentences… Functional syn-
tax, on the other hand, handles the way in which sentences are structured in terms
of smaller functional elements and eventually word.” (Morley 2000: 22). The argu-
ment that SFL lacks syntactic research put forward by Fawcett (2000) and Huang
(2000) refers to that of formal syntax, because functional grammar is mainly about
syntax. According to Huang (1999), textual analysis without grammatical analy-
sis is analysis not at all. In recent years, functional syntactic research has been
increasingly attracting the attention of systemicists, such as Matthiessen (1995),
Lock (1996), Morley (2000, 2004), Fawcett (2000, 2008, 2009), He and Gao
(2011), and Zhang (2012). Their researches have made considerable achievements,
and many traditional syntactic phenomena have been discussed in the framework
of SFL (e.g., Collins 1991; Tucker 1998; Huang 2003; Yang 2003, 2004; Zeng
2006). However, absolute clauses have not been touched upon, nor the term abso-
lute clauses has been referred to by systemicists, except that in discussing the
logico-semantic relations realized by non-finite clauses, Halliday points out that
“there may be an explicit Subject in the dependent clause” (Halliday 1994: 229).

1.2 Purpose of the Study

Traditional grammarians have conducted plenty of prescriptive researches on the


structures and functions of absolute clauses. This leaves us a large room for think-
ing: Can absolute clauses express other logical relationships except for those of
time, cause, condition, and concession? In addition to subordination, can abso-
lute clauses express coordination? SFL has developed a meaning-based system-
atic theory of functional syntax and does not prescribe language phenomena. As
to clauses, SFL does not confine the descriptions to such relations as time, cause,
condition, and concession.
The general purpose of this study, therefore, is to discuss absolute clauses in
the theoretical framework of SFL. The specific purpose is to look into the types of
4 1 Introduction

relationship absolute clauses can realize, and the synchronic and diachronic dis-
tributions of absolute clauses in actual language use. We wish that this could not
only make up for the deficiencies of traditional form-based grammatical research
on absolute clauses, but also make some contribution to SFL.

1.3 Organization of the Book

This research consists of three parts. The first part includes the introduction, the
literature review, the theoretical basis, and the research design. The second part is
the SFL research of absolute clauses. The third part is a corpus-based quantitative
research of absolute clauses and the conclusion.
The first part includes four chapters. This chapter is the introduction. Chapter 2
reviews the status quo of the research on absolute clauses in order to reveal the
problems that lie behind the current research. Chapter 3 puts forward the SFL
approach to absolute clauses, offering a sketch of the functional syntactic theory,
and proposes the identification criteria of absolute clauses through a SFL defini-
tion of absolute clauses and a discussion of the conditions of formation. Chapter 4
introduces the research questions, methodology, and the method of data collection
and analysis.
The second part is the fifth chapter. This chapter assumes the meaning potential
of absolute clauses according to the identification criteria proposed in Chap. 3 and
discusses the relationships realized by absolute clauses in the network of clause
complex. Different types of absolute clauses are also different in the tendency to
be independent from the primary clause. The independence of absolute clauses
is embodied on two dimensions, i.e., that of primary clauses and that of absolute
clauses.
The third part includes Chaps. 6 and 7, which is a corpus-based quantitative
study of absolute clauses. Chapter 6 is mainly a research on the relations of elabo-
ration, extension, and enhancement of expansion and that of projection realized by
absolute clauses, involving such aspects as functional distribution, stylistic distri-
bution, historical distribution, and case choice. Chapter 7 analyzes and discusses
the research results in Chap. 6.
Chapter 8 is the conclusion of this research. It summarizes this research, points
out its problems, and suggests future researches in this field.

References

Al-Hamash, K. I., & Abdulla, J. J. (1979). A course in modern English grammar. Baghdad:
IDELTI.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spo-
ken and written English. London: Longman.
Collins, P. C. (1991). Cleft and pseudo-cleft constructions in English. London: Routledge.
References 5

Crystal, D. (2008). A dictionary of linguistics and phonetics (6th ed.). Oxford: Blackwell
Publishing.
Curme, G. O. (1931). A grammar of the English language: Syntax. Boston: D.C. Heath.
Fawcett, R. P. (2000). Theory of syntax for systemic functional linguistics. Philadelphia,
Amsterdam: John Benjamins.
Fawcett, R. P. (2008). Invitation to systemic functional linguistics through the Cardiff gram-
mar: An extension and simplification of Halliday’s systemic functional grammar (3rd ed.).
London: Equinox.
Fawcett, R. P. (2009). Functional syntax handbook: Analyzing English at the level of form.
London: Equinox.
Fowler, H. W. (1965). A dictionary of modern English usage (2nd ed.). Oxford: Oxford
University Press.
Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward
Arnold.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (1999). Construing experience through meaning:
A language-based approach to cognition. London, New York: Cassell.
He, W., & Gao, S.-W. (2011). Studies of functional syntax. Beijing: Foreign Language Teaching
and Research Press.
Huang, G.-W. (1999). A functional approach to English syntactic analysis. Journal of Sun Yatsen
University (Social Science Edition), 4, 20–27.
Huang, G.-W. (2000). Systemic-functional linguistics: Forty years on. Foreign Language
Teaching and Research, 1, 15–21.
Huang, G.-W. (2003). Enhanced theme in English: Its structures and functions. Taiyuan: Shangxi
Education Publishing House.
Jespersen, O. (1949). A modern English grammar on historical principles. London: Allen &
Unwin.
Kruisinga, E. (1932). A handbook of present-day English part II. English accidence and syntax.
Groningen: P. Noordhoff.
Lock, G. (1996). Functional English grammar: An introduction to second language teachers.
Cambridge: CUP.
Matthiessen, C. M. I. M. (1995). Lexico-grammatical cartography: English systems. Tokyo:
International Language Sciences Publishers.
Morley, G. D. (2000). Syntax in functional grammar: An introduction to lexicogrammar in sys-
temic linguistics. London: Continuum.
Morley, G. D. (2004). Explorations in functional syntax: A new framework for lexicogrammatical
analysis. London: Equinox.
Morris, R. (1886[2010]). Historical outlines of English accidence: Comprising chapters on the
history and development of the language and on word formation. Charleston: BiblioBazaar.
Oliphant, T. L. K. (1878). The old and middle English. London: Macmillan.
Onions, C. T. (1905[2010]). An advanced English syntax. Whitefish: Nabu Press.
Pence, R. W., & Emery, D. W. (1965). A Grammar of present-day English (3rd ed.). New York:
The Macmillan Company.
Poutsma, H. (1929). A grammar of late modern English. Groningen: P. Noordhoff.
Quirk, R., & Greenbaum, S. (1973). A university grammar of English. Essex: Longman.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London, New York: Longman.
Ross, C. H. (1893). The absolute participle in middle and modern English. Publications of the
Modern Language Association of America, 8(3), 245–302.
Stump, G. T. (1985). The semantic variability of absolute constructions. Dordrecht: D. Reidel
Publishing Company.
Sweet, H. (1903). A new English grammar logical and historical part II: Syntax. Oxford:
Clarendon Press.
6 1 Introduction

Tucker, G. (1998). The lexicogrammar of adjectives: A systemic functional approach to lexis.


London: Cassell.
Visser, F Th. (1972). Historical syntax of the English language. Leiden: E.J. Brill.
Wooley, E. C. (1920). A handbook of composition: A compendium of rules. Boston: D.C. Health.
Yang, B.-J. (2003). A study of non-finite clauses in English: A systemic functional approach.
Beijing: Foreign Language Teaching and Research Press.
Yang, B.-J. (2004). Towards the criteria of non-finite clause identification: A systemic-functional
approach. Language Sciences, 26, 233–249.
Zeng, L. (2006). A functional interpretation of “projection”. Guangzhou: Sun Yat-sen University
Press.
Zhang, D.-L. (2012). Syntax in systemic functional linguistics. Journal of Tongji University
(Social Science Section), 1, 89–98.
Chapter 2
Absolute Clauses in the Literature

Absolute clauses form a category of absolute construction. “The term absolute


derives from Latin absolūtum, meaning ‘loosened from’ or ‘separated’” (Wheelock
and LaFleur 2005: 155–157), “standing out of the syntactic connexion” (Jespersen
1937: 126), “or abnormally connected to the rest of the sentence” (Crystal 2008: 2).
For example, the adverb or adjective at the beginning of a sentence in English is
an absolute constituent. See 2-1:
2-1a. However, it suffers from several disadvantages. (BNC_MISC)
b. Alone, Sara switched on her radio and did the washing-up. (BNC_FIC)
However, an absolute constituent does not necessarily constitute an absolute
construction, because an absolute clause has its own syntactic structure which is
unnecessary for an absolute constituent. Absolute clauses are “referred to as abla-
tive absolutes in Latin” (Wheelock and LaFleur 2005: 155), having conceptual
meaning but “not dependent on the subject of the main clause” (Zandvoort 1972:
37), and “are self-contained in idea and not grammatically tied to the sentence”
(Kane 1983: 754). They do not modify or connect with any single term but rather
modify the entire idea expressed in the main clauses. In 2-1, however and alone
are free from the clause in form. The former is an conjunctive adverb meaning
adversative, but an conjunctive word itself has not its own syntactic structure; the
latter can be seen as the reduction of a finite subordinate clause, or as non-finite
clause being alone with being omitted, hence an absolute construction.

2.1 What Is Absolute Clause

Traditional grammar considers such constructions consisting of a logical subject


and a logical predicate functioning as adverbials as nominative absolute clauses
or nominative absolute participles. It is nominative because the subject of this

© Springer-Verlag Berlin Heidelberg 2015 7


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_2
8 2  Absolute Clauses in the Literature

construction is always nominative, or at least the subject of the English absolute


clauses is nominative; it is participle because the logical predicate is always par-
ticipial. Jespersen (1937: 126) considers that the two names are both inappropri-
ate and consequently uses the term “nexus tertiary.” This is because the case in
absolute clauses may also be accusative (Curme 1931: 154; Jespersen 1937: 126;
Visser 1972: 1148); in addition to present participles, infinitives, past participles,
nouns, adjectives, adverbs, and prepositional phrases can also be used as the logi-
cal predicate. Among these types, “present participial clauses are the commonest”
(Al-Hamash and Abdulla 1968: 176, 1979: 376) and are the standard absolute con-
structions (Reuland 1983: 127). Fowler (1965: 4) also refers to this construction as
an absolute clause. However, absolute clauses defined by Fowler (1965) consist of
a noun or pronoun that is not the subject or object of any verb or the object of any
preposition in the main clauses. As we can see, the logical predicate of the abso-
lute clause in this sense is the non-finite verb, which is equivalent to the absolute
participle in the traditional sense, with verbless clauses not included. Non-finite
adverbial clauses or verbless adverbial clauses with explicit subjects are referred
to as absolute clauses (Quirk and Greenbaum 1973, Quirk et al. 1985; Watson
1976; Biber et al. 1999), and absolute clauses are not introduced by subordina-
tors (Quirk et al. 1985: 1120; McArthur 1992: 6). They are independent from the
main clauses in form, used to modify the rest of the sentence (Watson 1976: 758).
They are the reduced finite clauses, hence belonging to the category of subordinate
clauses (Curme 1931: 156–157; Quirk and Greenbaum 1973: 348–351; Shopen
1985: 200–201). The different names used by different grammarians are summa-
rized in Table 2.1.
Among these names of absolute clauses, the gerund and the participle types
are not appropriate, because the gerund includes only the -ing participles, thus
excluding all verbless constructions, past participles, and infinitives, and the par-
ticiple clauses consist of only the -ing verbs and the past participles, with the
verbless constructions and infinitives excluded. There are some grammarians
who do not distinguish free adjuncts from absolute clauses, referring to both as
absolutive clauses (e.g., Thompson and Longacre 1985) or gerundivized clauses
(e.g., Talmy 1978).
Some of these names define the case of absolute clauses, such as nominative
absolute construction, absolute nominatives, and nominative absolute; some oth-
ers define the syntactic functions of absolute clauses, such as absolute adjuncts,
nexus tertiary, and absolute free adjuncts. It can be seen that the names of absolute
clauses may have two uncertainties: The case of absolute clauses is not necessarily
nominative; the syntactic functions of absolute clauses are not necessarily adver-
bial. Nevertheless, there is one thing that is certain, namely “absolute.” “Absolute”
is manifested from two aspects: One is that the subject is not co-referential with
the subject of the main clause; the other is that it does not need a conjunctive
expression. However, the subject of absolute clauses and that of the main clauses
are sometimes co-referential. For example,
2.1  What Is Absolute Clause 9

Table 2.1  Names of absolute clauses


Grammarian Name of the construction
Annema (1924) Absolute structures
Annema (1924) Absolute constructions
Berent (1973, 1975)
McCawley (1983)
Mitchell (1985)
Stump (1985)
Berent (1973, 1975) Absolutes
Stump (1985)
Beukema (1980, 1982) Absolute free adjuncts
Kruisinga (1932)
Scheurweghs (1969)
Curme (1931) Absolute nominatives
Fowler (1860) Nominative absolute
Poutsma (1929)
Kruisinga (1932)
Reuland (1983)
Grady (1972) Simple absolutes
Fowler (1965) Absolute clauses
Watson (1976)
Haiman and Thompson (1984)
Quirk et al. (1985)
Biber et al. (1999)
Jespersen (1949) Nexus tertiary
Reuland (1983) Nominative absolute construction
Talmy (1978) Gerundivized clauses
Thompson and Longacre (1985) Absolutive clauses
Visser (1972) Absolute adjuncts
Zandvoort (1972) Absolute participle construction

2-2a. 
Our guest offering his assistance, he was accepted among the number.
(The Vicar of Wakefield, 1766)
b. 
The whole building being of wood, it seemed to carry every sound, like a
drum. (Women in Love, 1921)
The purpose of using absolute clauses is to avoid the subjects of the two clauses
referring to the same person or thing, and the sentences in 2-2 are rare and uncom-
mon (Onions 1905[2010]).
Latin ablative absolutes are usually translated as the “with + noun + partici-
ple” constructions in English. Influenced by ablative absolutes, traditional gram-
marians generally hold that absolute clauses are not introduced by a subordinate
conjunction, but they are always introduced by with. Here “introduced by with”
has two possible interpretations: One is that the logical subject-predicate construc-
tions following with are absolute clauses; the other is that the with constructions
themselves are absolute clauses.
10 2  Absolute Clauses in the Literature

According to the first interpretation, the preposition with is used to introduce


the subject of non-finite verbless clauses (Quirk et al. 1985: 993), and the con-
structions following with are still absolute clauses. However, in addition to with,
other prepositions can also introduce the subject of non-finite or verbless clauses,
as in 2-13. If the subject–predicate constructions introduced by the preposition
with are absolute clauses, then the other subject-predicate constructions in 2-13 are
also absolute clauses, which, however, are not included in the category of absolute
clauses by traditional grammarians.
The logical predicate of absolute clauses is the complement of the subject in
grammar, but in the logical subject-predicate constructions introduced by the prepo-
sition with, the logical subject is the object of with, and the logical predicate is the
complement of the object. However, the object–complement constructions are not at
all absolute. Jespersen (1933: 309) refers to this kind of construction as simple nexus,
which “may be the object not only of a verb, but also of a preposition…, as with
often means virtually the same thing as having” (Jespersen 1933: 312). The simple
nexus introduced by with has no direct grammatical relations with the main clause.
According to the second interpretation, the “with + noun + participle” construc-
tion is itself an absolute clause. In traditional grammar, with and the following simple
nexus together form a prepositional phrase, functioning as the adverbial of the sen-
tence, the same as nonfinite clauses headed by having. Since there is no conjunctive
expression, they can be referred to as free adjuncts, with the logical subject being the
subject of the main clauses. Therefore, the “with + noun + participle” constructions
can be considered as absolute clauses from neither of the two interpretations.

2.2 Types of Absolute Construction

Absolute constructions in traditional grammar are included in the grammatical


category of adverbial clauses, including free adjunct and nominative absolute. In
addition, Stump (1985: 1) subsumes a third type of absolute construction, aug-
mented absolute construction, headed by preposition with. For example,
2-3a. Standing up, she looked around the familiar room. (BNC_FIC) (free adjunct
construction)
b. Health permitting, her early retirement would be out of character.
(BNC_NA) (nominative absolute construction)
c. For months he had hung between life and death, with a bullet in his spine.
(BNC_FIC) (augmented absolute construction)

2.2.1 Free Adjunct Construction

Free adjunct constructions refer to the adverbial clauses with no explicit subject;
they are not connected with the main clauses with conjunctive expressions and are
usually separated by punctuation marks. The logical roles played by free adjuncts
2.2  Types of Absolute Construction 11

are diverse, such as time, cause, attendant circumstance, manner, result, condi-
tion, or concession (Frank 1972: 312–213; Zandvoort 1972: 37; Kane 1983: 756;
Stump 1985: 2). This construction “is native to English; examples can be found
in abundance from Old English through Modern English” (Stump 1985: 37).
Absolute adjuncts are structurally diverse, with the core components being non-
finite verbs, nouns, adjectives, adverbs, prepositional phrases, and so on, hence
nonfinite clauses in complex sentences. For example,
2-4a. 
Walking home, he goes through one large garden gate, only to see the other
one fall down. (BNC_MISC)
b. 
A carpenter, he had a long record of hard work, and his family was depend-
ent on his income. (BNC_NEWS)
c. 
Unable to meet his eyes, she stared at the garden, wondering vaguely why it
looked the same when she felt so very different. (BNC_FIC)
d. 
In his teens, he learned to drink and swore an allegiance to the pint.
(BNC_FIC)
In free adjuncts with non-finite verbs as the core components, the non-finite verbs
can be present participle, past participle, or infinitive. For example,
2-5a. Glancing up at the sky, she saw the storm clouds gathering on the horizon.
(BNC_FIC)
b. Published in 1941, it influenced many biographers. (BNC_MISC)
c. To tell you the truth, we are a bit at a loss. (BNC_FIC)
A free adjunct can also be embedded into another subordinate clause. See 2-6:
2-6 When the case, sitting at Chichester Rents in central London’s Chancery
Lane, opened nearly two months ago, Mr. Rook told the jury the fraud
involved 219 m. (BNC_NEWS)
The subjects of free adjunct constructions in the above example sentences are
all controlled by the subjects of the superordinate clauses, hence “related free
adjuncts” (Visser 1972: 1132). It is also possible that a free adjunct construc-
tion is not controlled by the subject of the superordinate clause; “in this case the
adjunct is termed unrelated or (more prescriptively) misrelated” (Stump 1985: 7),
or referred to as “dangling participle” (Visser 1972: 1132). For example:
2-7a. 
Crossing the road, a lorry knocked him down. (Bailie and Kitchen 1979: 294)
b. 
Playing backgammon and swapping jokes, the evening passed very pleasantly.
(Hodges and Whitten 1977: 254)
The free adjunct construction in 2-7a is syntactically related with a lorry, subject
of the main clause. In other words, grammatically, it is a lorry that is crossing the
road, but logically the subject of the free adjunct construction refers to him, object
of the main clause. In 2-7b, the free adjunct construction has no reference in the
main clause, that is, to say that the logical subject of the free adjunct construction
is implicit in the main clause, where there appears only the grammatical subject,
but this will not affect the effectiveness of communication.
12 2  Absolute Clauses in the Literature

Dangling construction has been used in all times throughout the history.
Grammarians use different names for this construction, such as loose partici-
ple (Jespersen 1949; Vallins 1952), fused participle or sentry participle (Fowler
1965: 215, 438), unattached participle, unrelated participle, pendant participle
or dangling participle (Quirk et al. 1985: 1121, 1123), and detached participle
or murky participle (Kane 1983: 255, 761). Grammarians have long been con-
cerned with dangling participles, such as Lowth (1762), Grant (1808), Cooper
(1831), and Curme (1912). Dangling participles are accepted as grammati-
cal without any criticism and “apparently not considered as irregular.” (Visser
1972: 1140).
It is Bain (1863) who first found fault in dangling construction. He “con-
demns the usage as an error arising from confounding the participle adjunct with
the absolute construction” (Visser 1972: 1140). Many other grammarians (e.g.,
Onions 1905[2010]; Partridge 1949; Fowler 1965) also question this construction,
for the logical subject and the grammatical subject of a dangling participle are not
co-referential. Although this may not cause misunderstanding, most grammarians
suggest avoiding this construction in language practice (e.g., Smart 1931; Hodges
and Whitten 1977; House and Harman 1950).
“Nowadays, in ‘literary’ English the idiom is avoided” (Visser 1972: 1140),
except occasionally appearing in premeditated spoken English, because although
it is ambiguous in structure, it is seldom misunderstood. Participles in some dan-
gling constructions have been being fixed with the gradual weakening of the
requirement of subjects, as in 2-8a, and some even are grammaticalized into prep-
ositions, as in 2-8b.
2-8a. Strictly speaking, I’ve no role here any more. (BNC_FIC)
b. Considering the interest rate blows, the market turned in a resilient perfor-
mance. (BNC_NEWS)
According to Quirk et al. (1985: 1122–1123) and Greenbaum (1996: 337–
338), the dangling construction is acceptable if it is a style disjunct that has the
speaker’s I as the understood subject, if the understood subject is a generic you,
we, or one or if it refers to the whole of the host clause, or in scientific usage,
if the understood subject refers to the I or we of the speakers or writers. For
example,
2-9a. Putting it mildly, you have caused us some inconvenience. (Quirk et al.
1985: 1122)
b. To check on the reliability of the first experiment, the experiment was repli-
cated with a second set of subjects. (ibid.: 1123)
c. Being Christmas, the government offices were closed. (ibid.: 1122)
d. Unknown to his closest advisers, he had secretly negotiated with an enemy
emissary. (ibid.: 1122)
2.2  Types of Absolute Construction 13

2.2.2 Nominative Absolute Construction

Absolute clauses are “nonfinite and verbless adverbial clauses that have an overt sub-
ject but are not introduced by a subordinator and are not the complement of a preposi-
tion” (Quirk et al. 1985: 1120); they are also known as “adverbial participle clauses
and adverbial verbless clauses” (Greenbaum 1996: 338). The people or thing conduct-
ing the action of the non-finite verb is not co-referential with the people or thing as the
subject of the main clause, so it is likely to be confused with a dangling construction.
The major difference between the two is that an absolute clause consists of a logical
subject and a logical predicate, while a dangling construction has no subject. Like free
adjuncts, absolute clauses are also various in structure, and their syntactical structures
can be distinguished from two dimensions, i.e., types of subject and types of predi-
cate. Modern grammarians generally accept that the subject of an absolute clause may
be a zero case noun or a nominative pronoun. They categorize the syntactic types of
absolute clauses mainly from the core component of the logical predicate, which can
be a non-finite verb or a verbless component. Non-finite verbs include present parti-
ciples, past participles, or infinitives; and verbless components include nouns, adjec-
tives, adverbs, or prepositional phrases. These structural types are shown in Fig. 2.1.
2-10a. She lay for a long while, the tears falling. (BNC_FIC)
b. Fish done, spuds on, she got out the asparagus and found they were plas-
tic! (BNC_MAG)
c. He also gave advance information about an evening meeting…and a one-
day conference…, both events to take place in London. (BNC_ACAD)
d. Her parents, Dad a solicitor and Mum a former Welsh international, are
watching with her brother as she beats Jennifer Santrock. (BNC_NEWS)
e. He dead, and you dying, he gave you the kiss of life. (BNC_SPOK)
f. Exhausted and confused, I came, cap in hand, busking for help and half
sang, half cried. (BNC_MISC)
g. Episode over, put it out of your mind. (BNC_FIC)
Grammarians generally hold that absolute clauses are not popularly used in
Modern English, “apart from a few stereotyped phrases, absolute clauses are for-
mal and infrequent” (Stump 1985: 10; Quirk et al. 1985: 1120). For example,
2-11a. Weather permitting, the big helicopters will place them between the flows
and the town. (BNC_NEWS)
b. All things considered, she would be better married. (BNC_FIC)

Nonfinite verb Verbless component

Present participle Past participle Infinitive Noun Adjective Adverb Preposition phrase

Fig. 2.1  Core component of predicate of absolute clauses (Kortmann 1991: 10)


14 2  Absolute Clauses in the Literature

c. There being only one way to get out of here, she went to find Felipe.
(BNC_FIC)
d. It being Ten o’clock, the debate stood adjourned. (BNC_MISC)

2.2.3 Augmented Absolute Construction

In Modern English, absolute clauses are usually not introduced by subordinators


(Visser 1972: 1158; 1271–1277; Quirk et al. 1985: 1120; McArthur 1992: 6), but
may be introduced by the subordinators with and without (Quirk and Greenbaum
1973: 723, 726; Quirk et al. 1985: 1003; Biber et al. 1999: 137), forming aug-
mented absolutes. For example,
2-12a. But for the next decade or more, nearly all orchestral and instrumental
music was issued in cut-down form, without any warning being given.
(BNC_MISC)
b. With time running out, they desperately need points to avoid relegation.
(BNC_NEWS)
Some other prepositions can also introduce absolute clauses, but not as popu-
larly used as with and without. For example,
2-13a. By women being open about sex, it made life much easier for men. (BNC_NA)
b. They gave us some time back afterward, because of the brain having an
operation. (BNC_NA)
c. Despite turnover being virtually flat at 13,242  m, Vallance said that BT
had been successful in controlling operating costs. (BNC_NA)
Free adjuncts can also be introduced by subordinators which can be a conjunc-
tion or a preposition, forming augmented free adjuncts. When a free adjunct con-
struction is introduced by a subordinating conjunction, the verbal element heading
the adjunct is participial (Curme 1931: 276; Jespersen 1949: 407); but when a free
adjunct construction is introduced by a preposition, the verbal element must be a
gerund (Stump 1985: 12). For example,
2-14a. After leaving the magazine, Caroline worked for a short time in the
Cardiff newsroom. (BNC_MISC)
b. On leaving school, he worked in a chemist’s shop. (BNC_MISC)

2.2.4 Differences and Similarities

The absolute constructions, i.e., free adjuncts, nominative absolutes, and aug-
mented absolutes, are all reduced finite clauses (Curme 1931: 156; Quirk and
Greenbaum 1973). The reason why they are called absolute construction is that
2.2  Types of Absolute Construction 15

there are two important similarities among them: All of them have their own tone
and are separated by a pause from the main clause, or by a comma in writing;
all of them are non-finite or verbless adverbial clauses, hence no tense or mood
marks.
However, there are also significant differences among them. Free adjuncts and
absolute clauses are not introduced by an explicit subordinator and are not depend-
ent on the main clause. This is the indicator of difference between these two and
augmented absolutes. Quirk et al. (1985: 1123) refer to these two types of absolute
constructions as supplementive clauses.
Moreover, there are two differences between free adjuncts and absolute clauses.
First, the subject of a free adjunct is implicit, and the implicit subject can be co-
referential with the subject of the main clause, hence related free adjunct or unre-
lated free adjunct. However, the subject of an absolute clause is explicit, and this
explicit subject is not co-referential with the subject of the main clause. Second,
free adjuncts can be introduced by a subordinator which can be a conjunction or
a preposition, while an absolute clause cannot be introduced by a conjunction, but
can by a preposition, such as with, forming an augmented absolute. From this per-
spective, a free adjunct is called absolute construction because there is not a sub-
ordinator connecting it with the main clause. An absolute clause is called absolute
construction, for it has not a subordinator to connect with the main clause, nor
a co-referential subject with the main clause. Free adjuncts are non-finite clauses
without explicit subordinators, and augmented absolutes are logical subject-predi-
cate constructions with explicit subordinators.

2.3 Logical Roles of Absolute Clauses

Traditional grammarians focus on the syntactic structures and logical roles of


absolute clauses in their studies. In this section, we will review the logical roles of
absolute clauses and their determiners.
Many traditional grammarians have carried out researches on the logical roles
expressed by free adjuncts and absolute clauses. For example, Curme (1931:
154–157) distinguishes the following six logical roles: time, cause, condition and
exception, attendant circumstance, manner proper, and concession. Since Curme
(ibid.: 158) considers free adjuncts as reduced absolute constructions, he does not
distinguish free adjuncts from absolute clauses in his classification of logical roles
of absolute clauses.
Kruisinga (1932: 274–275) subsumes all logical roles assumed by free adjuncts
under the notion of attendant circumstances, including the following four types:
(i) reason or cause: Shy, reserved and proud, I would have died rather than have
breathed a syllable of my secret.
(ii) difference of time: Finding Blanche determined, Father Andre presently took
his leave.
16 2  Absolute Clauses in the Literature

(iii) contrast: Like all craftsmen of the kind, he is at the mercy of his material.
(iv) alternative circumstances: Genuine, or a joke of the enemy, it spoke waken-
ing facts to him.
Although Kruisinga (1932: 280) has not listed the logical roles of absolute
clauses, he considers absolute clauses themselves as free adjuncts. Absolute
clauses differ from free adjuncts only in that they have their own subjects, so the
logical roles assumed by free adjuncts are the same as those assumed by absolute
clauses in his understanding.
Jespersen (1949: 61–64) thinks that it is not always easy or even impossible to
draw a clear line between several applications, but he suggests four logical roles of
absolute clauses:
(i) cause or reason: The wise men of antiquity…were afraid that—men being
what they are—their discoveries might be put to bad or futile uses.
(ii) condition: Conciliation failing, force remains; but force failing, no further
hope of reconciliation is left.
(iii) time: And the meal being over, he took Mr. Kaye into the other room.
(iv) descriptive circumstances: He remained in town, his idea being that he
wanted everything settled before his departure.
Visser (1972: 1054–1056; 1132–1139; 1149–1158; 1252–1255; 1266–1271)
discusses the range of logical roles played by free adjuncts and by absolute clauses
separately. He distinguishes four logical roles of absolute clauses: attendant cir-
cumstances; reason, ground, cause, or motive; time; and condition.
According to Quirk and Greenbaum (1972: 762), except for the attendant circum-
stances asserted by Kruisinga (1932), free adjuncts and absolute clauses have a more
specific sense in context. They suggest three logical roles played by absolute clauses:
(i) cause: All our savings gone, we started looking for jobs.
(ii) time: Cleared, this site will be very valuable.
(iii) circumstance: A case in both hands, Mabel stalked out of the house.
The above classifications of logical roles played by absolute clauses are repre-
sented in Table 2.2.
Adverbial clauses in traditional grammar can play the logical roles of time,
cause, place, condition, concession, manner, and comparison. However, there
are no adverbial clauses of attendant circumstance. Attendant circumstances are

Table 2.2  Logical roles of absolute clauses


Time Cause Circumstance Condition Concession Manner
Curme √ √ √ √ √ √
Kruisinga √ √ √ √
Jespersen √ √ √ √
Visser √ √ √ √
Quirk et al. √ √ √
2.3  Logical Roles of Absolute Clauses 17

usually introduced by prepositional phrases, and all can be introduced by with or


without (Quirk et al. 1985: 1003). For example,
2-15a. Without you to consult, I would be completely lost.
b. With the mortgage paid, they could afford to go abroad for their vacation.
c. Don’t walk around with your shirt hanging out.
d. With you as my friend, I don’t need enemies.
It should be noted that “grammars traditionally classify adjuncts on the basis
of meaning …rather than grammatical form, it is inevitably open-ended and the
boundaries between the different kinds are often quite fuzzy” (Huddleston and
Pullum 2005: 79). This is why grammarians distinguish types of logical roles
played by absolute clauses quite differently. The reason why the boundaries
between different logical roles are fuzzy is that absolute clauses do not need an
explicit coordinator. In fact, even a subordinating clause with an explicit subordi-
nator, no matter whether it is finite or nonfinite, may have multiple interpretations.
“They have the chameleon-like semantic quality of adapting to context” (Quirk
and Greenbaum 1972: 760). One conjunction can introduce subordinating clauses
of different semantic roles. For example, the conjunction since can introduce a
clause of time or reason. In addition, some clauses may be a combination of two
or more semantic roles. For example, some temporal clauses may imply relation-
ships of condition and concession, and some clauses of place may imply contrast
(Quirk et al. 1985: 1087).
Grammarians have also discussed the factors determining the syntactic logical
roles. According to Curme (1931: 155–156), the inversion of the subject and predi-
cate constituents occurs occasionally within absolute clauses expressing a causal,
conditional, circumstantial, or concessive relation, within which the concessive
relation appears more common. “This inversion is limited to a small number of
participles, some of which have taken on a prepositional value: except (excepted),
granted, given, during, pending, notwithstanding; or the value of a subordinating
conjunction, when the subject of the absolute is a clause: given that, granted that,
provided that” (Stump 1985: 19). See example 2-16:
2-16a. Given the School’s rapid rate of growth, more part-time tutors are needed
across the full range of management activities. (BNC_MISC) (cause)
b. Given sufficient advance notice, we’ll also be happy to carry collapsible
wheelchairs on any of our flights. (BNC_MISC) (condition)
c. Given the nomenclature of his position, he will prima facie be held out as
one of the members of the firm. (BNC_ACAD) (concession)
“The predicative constituent of a nominative absolute phrase expressing atten-
dant circumstances is commonly adverbial or prepositional” (ibid). For example,
2-17a. Richard went ahead, hands in pockets. (BNC_FIC)
b. He held out his hand to me, palm up. (BNC_MISC)
According to Jespersen (1949), the causal relationship is most usual with abso-
lute clauses having present participles, especially being in their predicates, as in
18 2  Absolute Clauses in the Literature

2-18a, and the relationship of time is most possible with absolute clauses having
past participles, adjectives, or adverbial phrases in their predicates, as in 2-18b–c.
Absolute clauses of time and condition usually precede the superordinate clauses,
as in 2-18b–d, and those expressing attendant circumstances are “generally
added after the main part of the sentence” (1949: 63), as in 2-18e–g. Like Curme
(1931), Jespersen (1949) also indicates that predicates of absolute clauses express-
ing attendant circumstances are usually adverbial or prepositional phrases, some
of which are, in Jespersen’s terms, the “condensed constructions”, such as hat in
hand, head first, and face down. See the following examples:
2-18a. The terrain being flat, the wind tore across scrub and heathland unim-
peded. (BNC_FIC) (cause)
b. Dishes done, I return home to find my bucket full of ‘slime’ which I pour
into the washing machine. (BNC_NA) (time)
c. Introductions over, Nicholson motioned for his guests to sit down.
(BNC_FIC) (time)
d. Weather permitting, the big helicopters will place them between the flows
and the town. (BNC_NEWS) (condition)
e. She rushed from the shop, hat in hand. (BNC_ACAD) (circumstance)
f. Then she dragged the body to the workbench and pushed it into the cup-
board, head first. (BNC_FIC) (circumstance)
g. Bissell’s body had been found lying on a pile of rubbish, face down.
(BNC_FIC) (circumstance)
“Both semantic and pragmatic factors may determine the logical role which
adjuncts and absolutes are felt to play” (Stump 1985: 22); Quirk and Greenbaum
(1972: 762) have discussed the pragmatic factors determining the logical roles of
absolute clauses. For example, the absolute clause in 2-19a may express the rela-
tionship of cause or time and that in 2-19b, condition or cause. “For the reader or
hearer, the actual nature of the accompanying circumstance has to be inferred from
the context” (Quirk et al. 1985: 1124). For example:
2-19a. Vanity overcoming discretion, Sherman phoned the Newark Evening News
to boast of his own treasure trove. (COHA_MAG)
b. Such being the case, a few remarks will be made on each kind separately.
(COHA_NF)
Quirk and Greenbaum (1972) also offer an explanation of semantic correlate of
absolute clauses. For example, “in—ing clauses, dynamic verbs typically suggest a
temporal link, and stative verbs a causal link” (Quirk and Greenbaum 1972: 762).
This can be taken as the third factor, the semantic factor that determines the logi-
cal relationship between the absolute clause and the main clause. For example,
2-20a. Money being scarce, Belen’s people don’t buy but barter. (COCA_FIC)
(cause)
b. Sanity returning, he ran after her. (BNC_FIC) (time)
2.3  Logical Roles of Absolute Clauses 19

Syntactic, pragmatic (reasoning), and semantic factors together determine the logi-
cal roles of absolute clauses. These factors help readers make a determination on
the most appropriate role from several possible logical roles.
Logical roles distinguished by grammarians can be classified into two catego-
ries: adverbial clauses and attendant circumstances. The difference between the
two is that the former can be expanded into clausal adjuncts introduced by sub-
ordinating conjunctions and the latter, prepositional phrases introduced by with or
coordinating clauses linked by conjunction and. For example,
2-21a. Dexter turned to Emma, eyes blazing an apology. (CLOB_P)
(circumstance)
b. There being no bridge, the master had to stop at the shore. (BROWN_E)
(cause)
c. This done, she contemplated with dismay the solitary hours that lay before
her. (LOB_P) (time)
d. The unexpected weather aside, it had been a good day. (CLOB_N)
(concession)
e. All things considered, the highway commissioners would seem to be
elected. (BROWN_C) (condition)
The absolute clause in 2-21a is an attendant circumstance, and those in 2-21b–e
play the relations of cause, time, concession, and condition, respectively. Due to
the absence of explicit conjunctive expressions, the logical roles expressed by
absolute clauses are always fuzzy. Grammarians such as Curme (1931), Jespersen
(1949), and Quirk and Greenbaum (1972, Quirk et al. 1985) have discussed on
the factors determining the logical relations assumed by absolute clauses, but clear
distinctions can hardly be attained in many cases. For example,
2-22a. Her fears somewhat lulled, she began to read. (LOB_N)
b. It was a very English sort of day, the air still, the sky a uniform white.
(CLOB_L)
c. The family circle was a tight one, the discipline strict. (FLOB_G)
d. That’s twice he did it, twenty years apart, two pregnancies ending in noth-
ing, nothing. (FLOB_K)
The absolute clause in 2-22a can be interpreted as expressing the relation of
time or cause. No matter which of the two relationships it expresses, it belongs
to the category of adverbial clauses. The distinctions between different types of
logical roles are also fuzzy. For example, the absolute clause in 2-22b can be con-
sidered as an adverbial clause expressing the relationship of cause or as an atten-
dant circumstance. It is grammatically acceptable that the preposition with can be
added to such absolute clauses as time, cause, and condition to form augmented
adjuncts. Some absolute clauses may have more interpretations, as in 2-22c, the
absolute clause can be taken as an adverbial clause of cause or an attendant cir-
cumstance or neither of the two. Rather, it is the explanation to the main clause.
Although traditional grammar does not distinguish the explanation type of abso-
lute clauses, the absolute clause in 2-22d is hard to be included in the categories
20 2  Absolute Clauses in the Literature

of clausal adjuncts or attendant circumstances. In fact, the absolute clauses in 2-23


“uniformly resist paraphrase by means of any sort of adverbial clauses” (Stump
1985: 334); they “are understood as explaining some notion ancillary to the mean-
ing of the main clauses” (ibid.: 335).
2-23a. …the orchestra played some blues, a gospel piece, and “I got Rhythm,”
sometimes in unison, sometimes in harmony, and sometimes free. (Stump
1985: 334)
b. The tombstones were spaced out on the floor in long rows, each stone
about six feet long and covered with carving in relief. (ibid.: 334)
c. To confront another person with one’s head uncovered was a grave insult,
the only exception being Taoist recluses and Buddhist priests. (ibid.: 334)
Most absolute clauses (about 80 %) are not typical adverbial clauses, but
appositive or coordinate clauses (Kortmann 1991: 99). Appositive (or coordinate)
absolute clauses can be considered as the explanation of the main clause or part of
the main clause, as in example 2-24:
2-24a. The city was alive in daylight, the Elves busy at their work, the streets
bustling with activity. (FROWN_N)
b. The MacGregors were there, Tim lounging on the piano stool, Susan
on the floor, Mrs. MacGregor upright and expressionless on the sofa.
(LOB_L)
The two absolute clauses in 2-24a are the explanations of the main clause,
while the three in 2-24b are the explanations of the nominal group The
MacGregors. There is a clear difference between these absolute clauses and the
absolute clause eyes blazing an apology in 2-21a. The absolute clause in 2-21a is
an attendant circumstance of the verb in the main clause and can be augmented
with a preposition with to form a prepositional phrase, and those in 2-24 are not
attendant circumstance because both verbs in the main clauses are not action
verbs, but linking verbs which cannot introduce attendant circumstances. Even if
the verb in the main clause is an action verb, the absolute clauses in 2-25 cannot
be considered as attendant circumstances either.
2-25a. There they continued their studies at the university, she in art, he in archi-
tecture. (BROWN_A)
b. For nearly forty years she and Sir Edwin Lutyens worked together—he
as architect of the house, she designing the garden—culminating in the
Viceroy’s House in New Delhi. (FLOB_G)
The predicate constituents in the two absolute clauses in 2-25a are a further
explanation of the main clause, which can be seen as finite clauses with continued
her study and continued his study omitted. The first absolute clause in 2-25b can
also be seen as a clause with worked omitted, and the predicate designing in the
second absolute clause is a hyponym of worked. Therefore, all the four absolute
clauses are not actions or state of affairs occurring accompanying the verbs of the
2.3  Logical Roles of Absolute Clauses 21

main clauses. There is a clear distinction between 2-25 and 2-26 although they
appear the same in structure.
2-26 The Benbergs stood by, he clasping his hands and watching her closely,
she wiping a plate round and round with a sodden cloth. (Visser 1972)
According to Visser (1972), the absolute clauses in 2-26 are attendant circum-
stances. In fact, the relationships expressed by these two absolute clauses are also
to some extent fuzzy. From the perspective of predicate, the verbs in the two abso-
lute clauses are both accompanying the verb stood in the main clause. However,
from the perspective of subject, the two absolute clauses can both be seen as the
explanation of the nominal group the Benbergs, subject of the main clause. Even
so, if the main verb stood is interpreted as a state of affairs, as The Benbergs were
there, the two absolute clauses are more inclined to function as explanation.

2.4 The Case of Absolute Clauses

There are two different interpretations for the case of absolute clauses to change
from dative to nominative. One believes that absolute clauses originated from
dative absolutes in the Old English, which were borrowed from Latin ablative
absolutes. According to this interpretation, the dative case in Anglo-Saxon is the
origin of the absolute clauses in English, and the absolute case changed from
dative to nominative owing to the loss of case inflections. “The inflections having
decayed, the dative was mistaken for the nominative” (Kellner 1892: 125). On the
change of absolute case from dative to nominative, Bright (1890: 159–162) wrote
the following:
Let us look at the history of the absolute construction in English. We begin with the
dative absolute in Anglo-Saxon (in origin a translation of the Latin ablative absolute); as
inflections break down we come upon the transition or ‘crude’ type, in which the pro-
noun remains dative in form while the participle has lost all signs of inflection. But all
nouns, as well as the participle, came to lose the inflectional sings of the dative case; we
then obtained the ‘crude’ type, in which both noun and participle, though absolute, were
without any trace of inflection. The final act in this history was the admission of the nom-
inative forms of the personal pronouns into this crude absolute construction – a dative
absolute in disguise.

Another interpretation believes that the use of a noun in the zero form or a
pronoun in the subject form is “a continuation of the Old English usage with the
noun before the participle in the zero case, with later analogous introduction of
the subject form of the pronouns. This latter phenomenon took probable place
as early as the middle of the fourteenth century” (Visser 1972: 1149). According
to this interpretation, the subject of absolute clauses should always be nomina-
tive because they evolved from the Old English. An absolute clause “is errone-
ous in making it the objective” (Murray 1808[2011]: 201), and Lowth (1762: 116)
warned against dative absolute in disguise “forcing of the English under the rules
22 2  Absolute Clauses in the Literature

of a foreign language”. “A Noun or a Pronoun is put absolute in the nominative,


when its case depends on no other word” (Brown 1861), hence called nominative
absolute (Fowler 1860: 517; Tipping 1961: 184).
However, due to the impact of the Latin grammar, many grammarians in the
nineteenth century held that it is not correct for zero case nouns or nominative
pronouns to be the subject of absolute clauses and insisted that the zero case
nouns or nominative pronouns before the absolute clauses are dative or ablative.
For example, in Modern English as well as in the Old English, absolute words
are always in the dative case. “The meaning conveyed by these absolute words
cannot be expressed by a true nominative” (Adams 1858[2010]). Pronouns can-
not change the characteristics of absolute clauses. Despite the use of the nomi-
native forms of the personal pronouns, absolute clauses are “historically the
objective absolute” (Bright 1890: 161). “The ablative absolute may be translated
by the English objective absolute, which is a close equivalent” (Gildersleeve
1888: 137–157). Although the nominative has taken the place of the dative, “yet
it is right to parse the so-called nominative absolute as ‘a dative absolute in dis-
guise’” (Ross 1893: 294).
Grammarians in the twentieth century generally accepted that absolute clauses
are nominative (e.g., Jespersen 1933), but there are still grammarians who believed
that absolute clauses can also be accusative (e.g., Curme 1931). Such clauses
include only a small number, some of which are obviously affected by Latin.
However, “among speakers of standard English, absolutes whose subjects are
oblique in case are generally regarded as unacceptable” (Stump 1985: 11). Fowler
(1965: 4) tries to explain that absolute clauses should be nominative by using
example 2-27.
2-27 There being no evidence against him, and he (not him) denying the
charge, we could do nothing. (Fowler 1965: 4)
Recent grammarians (e.g., Quirk et al. 1985; Biber et al. 1999; Crystal 2008,
etc.) not only think that absolute clauses are nominative, but also accept that some-
times absolute clauses can be accusative. Therefore, they use the term “absolute
clauses” instead of “absolute nominative clauses” to avoid the case problem to
some extent.
We are not sure whether absolute clauses (e.g., he liked) have evolved from the
dative absolute (e.g., him likade) of the Old English, or we should take them as a
continuation of the Old English usage with the noun before the participle in the
zero case. Although many grammarians of the nineteenth century held that it is not
correct for zero case nouns or nominative pronouns to be the subject of absolute
clauses, there are many absolute clauses with nominative pronouns as subject in
the works of this period. For example,
2-28a. There would he kneel to me in the snow…, he shivering with cold, and I
with apprehension. (1774)
b. Meanwhile both suffered, she not knowing why. (1888)
2.4  The Case of Absolute Clauses 23

c. With a new formality and silence she led the way into the hall, he following.
(1894)
d. She had turned back to the drawing-room, forgetting the other guests, he
walking beside her. (1894)
e. The Benbergs stood by, he clasping his hands and watching her closely, she
wiping a plate round and round with a sodden cloth. (1894)
f. She being down, I have the placing of the British crown. (1894)
It is noteworthy that accusative pronouns as subject of absolute clauses com-
pletely disappeared after the fifteenth century and reappeared in informal English
in the nineteenth century (Visser 1972: 1147). There are two possible interpreta-
tions: One is that absolute clauses with accusative pronouns as subject existed in
spoken English in this period of time; the other is that the use of accusative pro-
nouns in Modern English is the same as that of such expressions as it is me, that’s
him, etc. in origination. According to Burn (1766[2010]: 61), despite the fact that
2-29 is not correct in grammar, it is still necessary to speak like that.
2-29 Him watching, all the rest went to repose themselves. (1766)
Bain (1904: 273) found that the accusative form corresponding to the dative
form in the oldest English is not unusual until recently. For example,
2-30a. But you see, him being here, in the room—I had to be careful. (1926)
b. It made me so tired, it did. Him worshipping the ground she trod and her
not caring a snap of the fingers for him. (1932)
c. You’ve had a disappointment, I Know, her being away. (1933)
Grammarians have paid full attention to the case of absolute clauses, but they
have not reached any agreement so far. In actual language use, absolute clauses
can be nominative, as in 2-31, or accusative, as in 2-32.
2-31a. Of course he was thirteen years older than her, she being but twenty.
(BNC_FIC)
b. Some twenty thousand people attended to hear him speak, I being one of
them. (BNC_MISC)
c. He whispering endearments in his lover’s ear, the joy of lying in each oth-
er’s arms… (BNC_FIC)
2-32a. She was surprised he lived so poorly, him being a successful man.
(BNC_FIC)
b. Boys pouring into the room below, laughing, chattering, me seeing them
through the crack. (BNC_FIC)
c. You asked her if she would like to have the bairn, her being a minister’s
wife. (BNC_FIC)
24 2  Absolute Clauses in the Literature

2.5 Stylistic Effects of Absolute Clauses

Gildersleeve (1888: 137–157) studied the stylistic features of absolute clauses in


Greek. Inspired by Gildersleeve, Callaway (1889: 46–51) studied the stylistic fea-
tures of the Anglo-Saxon absolute clauses. The result shows that the stylistic effect
of the absolute clauses in Anglo-Saxon is much the same as that of Greek:
It gave movement to the sentence; it made possible flexibility and compactness. But,
owing to the artificial position of the absolute construction in Anglo-Saxon, its stylistic
value was reduced to a minimum, was indeed scarcely felt at all. The absolute partici-
ple rejected as an instrument of style, the Anglo-Saxon had no adequate substitute there-
fore. The two commonest substitutes, the dependent sentence and the co-ordinate clause,
as used in Anglo-Saxon, became unwieldy and monotonous. Brevity and compactness
were impossible; the sentence was slow in movement and somewhat cumbersome. The
language stood in sore need of a more flexible instrument for the notation of subordinate
conceptions, of such an instrument as the absolute dative seemed capable of becoming but
never became.

In the first half of the Middle Ages, absolute clauses were practically non-existent.
As Ross (1893: 296) states:
Its prevalence in Chaucer is due largely to Italian influence, in part also to French influ-
ence, and the occurrence of the participle in the works of Chaucer’s contemporaries and of
the fifteenth century writers is to be traced to the same French influence. But the construc-
tion was avoided as much as possible, and in its stead the various shifts that were resorted
to in Anglo-Saxon were used. The absolute participle here cannot be spoken of as “a norm
of style”… During the fifteenth century, however, just before the awakening caused by the
Revival of Learning, the absolute participle became, as we have seen, somewhat prevalent
and was more felt in the style. (Ross 1893: 296)

In the Modern English period, absolute clauses have assimilated and developed
into a style and were used by all writers. In fact, early in the second half of the
seventeenth century, absolute clauses became completely natural. They became
thoroughly fixed as a style in the eighteenth century. At this time, the appearance
of the novel as a style makes the use of absolute clauses very popular.
Absolute clauses are used differently in different styles of works. “In the six-
teenth and seventeenth centuries it belonged largely to didactic and philosophi-
cal prose, but now its province is distinctively narration and description…Next
to this stand biography, history, and the essay” (Ross 1893: 297–298). In the
Anglo-Saxon and medieval English, absolute clauses were nearly used only in
prose, rarely in poetry. Chaucer is an exception. This is because Chaucer imi-
tated Boccaccio in whose poems there are a large number of absolute clauses.
In Shakespeare’s poems absolute clauses are rarely seen, but very popular in the
poems of Dryden and the poets of the eighteenth and nineteenth centuries.
Absolute clauses were severely inveighed by grammarians of the nineteenth
century. “Even such forms as ‘Herod being dead, the angel warned Joseph’ seem
rare in the best recent English” (Mcelroy 1885: 105). “The participial construction
is a convenient means of condensation… Being, however, a subordinated construc-
tion, it needs careful adjustment to the principal assertion on which it depends”
2.5  Stylistic Effects of Absolute Clauses 25

(Genung 1885[2010]: 117). “This construction belongs to literary style rather than
to colloquial speech” (Jespersen 1933: 313).
In general, absolute clauses are a good choice which can give sentences life and
movement and make sentences free and diverse. “Because nonfinite clauses lack
tense markers and modal auxiliaries and frequently lack a subject and a subordi-
nating conjunction, they are valuable as a means of syntactic compression. Certain
kinds of nonfinite clause are particularly favoured in written prose” (Quirk et al.
1985: 995). In Modern English, absolute clauses have become an important part of
English syntax, forming a particular style. They are used in the works of all writ-
ers, both in literary works and in oral conversation, even though mainly confined
to some fixed expressions, such as all things being equal and all things consid-
ered. However, Modern grammarians (e.g., Quirk et al. 1985) generally hold that
absolute clauses are normal and infrequent, mainly used in formal works and sel-
dom in spoken or informal texts. Jespersen (1949: 62) illustrates the situations of
application of absolute clauses through 2-33 as follows:
2-33a. He stood, hat in hand.
b. He stood, his hat in his hand.
c. He stood, with his hat in his hand.
Obviously, a search of similar phenomenon in corpora reveals that 2-33b is
rarely seen, 2-33c is commonly used in spoken language, while 2-33a is relatively
common in literary works.

2.6 Questions to Be Answered

In spite of the subject–predicate structure, absolute clauses are not sentences in


the true sense. They are “so termed because they are not explicitly bound to the
matrix clause syntactically” (Quirk et al. 1985: 1120). In other words, absolute
clauses do not need to have a conjunction or preposition to express the relationship
between absolute clauses and the main clauses. However, as subordinate clauses,
the relationship does exist. Although with is an explicit conjunctive expression,
it cannot make explicit the implicit logical roles of absolute clauses. That is why
augmented absolutes belong to the category of absolute construction, and all non-
finite clauses with subject “except for that of the bare infinitive clauses may be
introduced by the subordinators with and without” (ibid.: 1003). However, gram-
marians have not explained why the absolute clauses in 2-34 cannot be introduced
by with.
2-34 There’ll be these terrible noises coming out of the woods, cars crashing,
elephants screaming. (COHA_MAG)
According to Kortmann (1991), the absolute clauses in 2-34 do not function
as adverbials, but as appositives; they express the relationship of explanation.
Adverbial clauses and appositive clauses are both subordinate clauses, but they are
26 2  Absolute Clauses in the Literature

different from the main clauses in position. Quirk et al. (1985: 1123) distinguish
the positions of attributive clauses and adverbial clauses with their main clauses.
2-35a. Jason, told of his son’s accident, immediately phoned the hospital. (Quirk
et al. 1985: 1123)
b. Jason, who was told of his son’s accident, immediately phoned the hospi-
tal. (ibid)
The position of adverbial clauses is arbitrary. They can be positioned ini-
tially, medially, and finally, while the most typical position of attributive clauses
is immediately after their antecedent. If subjectless nonfinite clauses occur in that
position, they may be indistinguishable from the participle clauses functioning as
post-modifiers or noun phrases in apposition. See 2-36 below:
2-36a. This substance, discovered almost by accident, has revolutionized
medicine.
b. This substance, which was discovered almost by accident, has revolution-
ized medicine.
c. Discovered almost by accident, this substance has revolutionized
medicine.
The non-finite clause in 2-36a can be interpreted as a post-modifier, as in
2-36b, or as a subjectless non-finite adverbial clause, as in 2-36c. The absolute
clauses in 2-34 are positioned after the antecedent and they cannot be positioned
before the main clause, hence not adverbial clauses but appositive clauses. The
same is true for the absolute clauses in 2-26.
According to the analysis above, it is problematic to define absolute clauses as
non-finite and verbless adverbial clauses with an explicit subject but without an
introducing subordinator, because being adverbial is not the necessary require-
ment for constituting absolute clauses. Absolute clauses are not always non-finite
or verbless adverbial clauses with an explicit subject; they may also be non-finite
or verbless appositive clauses with an explicit subject. Accordingly, at least three
questions in research available need to be answered.
Question One Since being adverbial is not the necessary requirement for form-
ing absolute clauses, then what are the identification criteria for
absolute clauses?
Question Two Since appositive clauses with an explicit subject can form abso-
lute clauses, then can other function types of nonfinite or verb-
less clauses with an explicit subject form absolute clauses?
Question Three What are the characteristics of historical, stylistic and case dis-
tribution of absolute clauses?
Many grammarians believe that absolute clauses are formal in style and are
decreasing in actual use, but the stylistic distributions of absolute clauses or the
function types of absolute clauses have not been explored in detail. Scholars dis-
tinguish different logical roles played by absolute clauses, but many problems still
remain. Deficiencies of current studies are obvious. In SFL, absolute clauses are
2.6  Questions to Be Answered 27

almost totally ignored too. Only when discussing non-finite clauses are absolute
clauses mentioned, i.e., “there may be an explicit Subject in the dependent clause”
(Halliday 1994: 229; Halliday and Matthiessen 2004: 404). See 2-37 below:
2-37a. John went off by himself, the rest of us staying behind.
b. It’s a much bigger house, for the children to have their own rooms.
We may simply say that, the non-finite clause in 2-37a, the rest of us staying
behind, has an explicit subject which does not refer to the subject of the primary
clause, hence is an absolute clause; and the nonfinite clause in 2-37b, for the chil-
dren to have their own rooms, has also an explicit subject, but it has an explicit
conjunctive preposition, hence is not an absolute clause. This is far from SFL anal-
ysis. More importantly, it is not readily applicable to many other instances.
In the next chapter, we will first offer a sketch of systemic functional theory,
and then discuss the SFL approach to absolute clauses.

References

Adams, E. (1858[2010]). The elements of the English language. Whitefish: Kessinger Publishing.
Al-Hamash, K. I., & Abdulla, J. J. (1968). A course in English linguistics. Baghdad: Al-Sha’b
Press.
Al-Hamash, K. I., & Abdulla, J. J. (1979). A course in modern English grammar. Baghdad:
IDELTI.
Annema, H. (1924). Die sogenannten absoluten participalkonstructionen im neuhochdeutschen.
Groningen: de Waal.
Bailie, J., & Kitchin, M. (1979). The Hamlyn guide to English usage. London: Hamlyn.
Bain, A. (1863). An English grammar. London: Longman.
Bain, A. (1904). A higher English grammar. London: Longman.
Berent, G. B. (1973). Absolute constructions as “subordinate clauses”. In C. Corum, C. Smith-
Stark, & A. Weiser (Eds.), You take the high node and I’ll take the low node: Papers from the
comparative syntax festival (pp. 147–154). Chicago: Chicago Linguistics Society.
Berent, G. B. (1975). English absolutes in functional perspective. In R. E. Grossman, L. J. San,
& T. J. Vance (Eds.), Papers from the parasession on functionalism (pp. 10–23). Chicago:
Chicago Linguistics Society.
Beukema, F. H. (1980). Prepositions and tenseless sentence adverbials in English. In S. Daalder
& M. Gerritsen (Eds.), Linguistics in the Netherlands 1980 (pp. 69–79). Amsterdam, Oxford
& New York: North-Holland.
Beukema, F. H. (1982). On the internal structure of free adjuncts. In S. Daalder & M. Gerritsen
(Eds.), Linguistics in the Netherlands 1982 (pp. 71–82). Amsterdam: North-Holland.
Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spo-
ken and written English. London: Longman.
Bright, J. W. (1890). The objective absolute in English. Modern Language Notes, 3, 159–162.
Brown, G. (1861). The grammar of Englsih grammars (6th ed). New York.
Burn, J. (1766[2010]). A practical grammar of the English language. Charleston: BiblioBazaar.
Callaway, M. (1889). The absolute participle in Anglo-Saxon. Baltimore: Press of I. Friedenwald.
Cooper, J. G. (1831). A plain and practical English grammar. Philadelphia: Judah Dobson.
Crystal, D. (2008). A dictionary of linguistics and phonetics (6th ed.). Oxford: Blackwell
Publishing.
Curme, G. O. (1912). A history of the English relative construction. London: Longman.
Curme, G. O. (1931). A grammar of the English language: Syntax. Boston: D. C. Heath.
28 2  Absolute Clauses in the Literature

Fowler, W. C. (1860). English grammar. New York: A.S. Barnes.


Fowler, H. W. (1965). A dictionary of modern English usage (2nd ed.). Oxford: Oxford
University Press.
Frank, M. (1972). Modern English: A practical reference guide. New Jersey: Prentice-Hall.
Genung, J. F. (1885[2010]). The practical elements of rhetoric: With illustrative examples.
Charleston: BiblioBazaar.
Gildersleeve, B. L. (1888). On the stylistic effect of the Greek participle. The American Journal
of Philosophy, 9(2), 137–157.
Grady, M. (1972). The English absolute construction. Linguistics, 90, 5–10.
Grant, J. (1808). Institutes of Latin grammar. London: Longman.
Greenbaum, S. (1996). The Oxford English grammar. Oxford: Oxford University Press.
Haiman, J., & Thompson, S. A. (1984). Subordination in universal grammar. In C. Brugman, &
M. Macaulay (Eds.), Proceedings of the tenth annual meeting of the Berkeley linguistics soci-
ety (pp. 510–523). Berkeley: Berkeley Linguistics Society.
Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward
Arnold.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2004). An introduction to functional grammar
(3rd ed.). London: Edward Arnold.
Hodges, J., & Whitten, M. E. (1977). Harbrace college handbook. New York: Harcourt Brace
Jovanovich.
House, H. C., & Harman, S. E. (1950). Descriptive English grammar. Englewood Cliffs:
Prentice-Hall.
Huddleston, R., & Pullum, G. K. (2005). A student’s introduction to English grammar.
Cambridge: Cambridge University Press.
Jespersen, O. (1933). Essentials of English grammar. London: Allen & Unwin.
Jespersen, O. (1937). Analytic syntax. London: Allen & Unwin.
Jespersen, O. (1949). A modern English grammar on historical principles. London: Allen &
Unwin.
Kane, T. C. (1983). The Oxford guide to writing: A rhetoric and handbook for college students.
New York: Oxford University Press.
Kellner, L. (1892). Historical outlines of English syntax. London: Macmillan.
Kortmann, B. (1991). Free adjuncts and absolutes in English: Problems of control and interpre-
tation. London: Routledge.
Kruisinga, E. (1932). A handbook of present-day English part II. English accidence and syntax.
Groningen: P. Noordhoff.
Lowth, R. (1762). A short introduction to English grammar. London: A. Millar and R. and J.
Dodsley.
McArthur, T. (1992). The oxford companion to the English language. New York: Oxford
University Press.
McCawley, J. K. (1983). What’s with with? Language, 59, 271–287.
McElroy, J. G. R. (1885). The structure of English prose: A manual of composition and rhetoric.
New York: Armstrong & Son.
Mitchell, B. (1985). Old English syntax (Vol. 2). Oxford: Clarendon Press.
Murray, L. (1808[2011]). An English grammar: Comprehending the principles and rules of the
language. Whitefish: Nabu Press.
Onions, C. T. (1905[2010]). An advanced English syntax. Whitefish: Nabu Press.
Partridge, E. (1949). English: A course for human beings. London: Winchester Publications.
Poutsma, H. (1929). A grammar of late modern English. Groningen: P. Noordhoff.
Quirk, R., & Greenbaum, S. (1972). A grammar of contemporary English. London & New York:
Longman.
Quirk, R., & Greenbaum, S. (1973). A university grammar of English. Essex: Longman.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London & New York: Longman.
References 29

Reuland, E. J. (1983). Governing-ing. Linguistic Inquiry, 14, 101–136.


Ross, C. H. (1893). The absolute participle in middle and modern English. Publications of the
Modern Language Association of America, 8(3), 245–302.
Scheurweghs, G. (1969). Present-day English syntax: A survey of sentence patterns. London:
Lowe & Brydone.
Shopen, T. (1985). Language typology and syntactic description. Cambridge: Cambridge
University Press.
Smart, W. K. (1931). Handbook of effective writing. New York & London: Harper & Brothers.
Stump, G. T. (1985). The semantic variability of absolute constructions. Dordrecht: D. Reidel
Publishing Company.
Talmy, L. (1978). Figure and ground in complex sentences. In J. H. Greenberg (Ed.), Universals
of human language (Vol. 4, pp. 625–649). Syntax Stanford: Stanford University Press.
Thompson, S. A., & Langacre, R. E. (1985). Adverbial clauses. In T. Shopen (Ed.), Language
typology and syntactic description (Vol. 2). Complex constructions Cambridge (pp. 171–234):
Cambridge University Press.
Tipping, L. (1961). Matriculation English grammar of modern usage. London: Macmillan.
Vallins, G. H. (1952). Good English: How to write it. London: A. Deutsch.
Visser, F Th. (1972). Historical syntax of the English language. Leiden: E. J. Brill.
Watson, O. (1976). Longman modern English dictionary (2nd ed.). London: Longman.
Wheelock, F. M., & LaFleur, R. A. (2005). Wheelock’s latin (6th ed.). New York: HarperCollins.
Zandvoort, R. W. (1972). A handbook of English grammar (6th ed.). London: Longman.
Chapter 3
Approaching Absolute Clauses
from the SFL Perspective

3.1 Introduction

From “Categories of the Theory of Grammar” (Halliday 1961[2002]) to Introduction


to Functional Grammar (Halliday 1985, 1994; Halliday and Matthiessen 2004,
2014), SFL created by Halliday has developed into a general linguistics from a
grammar theory and has become a mainstream linguistics in the world (Fawcett
2000: xviii; Eggins 2004: xiii). In “Categories of the Theory of Grammar” (Halliday
1961[2002]), Halliday outlined four categories of language: unit, structure, class, and
system. Later, in “Some Notes on ‘Deep’ Grammar” (Halliday 1966), he focused on
the two categories: system and structure. From the system category, he developed
systemic grammar, and from the structure category, he developed functional gram-
mar. In systemic grammar, language is considered as a meaning system, and form
is the realization of meaning. However, without a formal theory, the meaning will
not be realized. In functional grammar, form is represented as the constituent struc-
ture of rank, based on which the functional syntactic structures related to the three
metafunctions are established, including the transitivity structure realizing ideational
metafunction, the mood structure realizing interpersonal metafunction, and the the-
matic and information structures realizing textual metafunction. Therefore, “syntax
plays a very important part, even a central part in SFL theory” (Huang 2007).
Functional syntax emphasizes the meaning realization of form and takes clause
as the basic analyzing unit. After analyzing the functional structures of clauses,
functional grammar analyzes the functional structures of the units below the clause
(group/phrase), above the clause (clause complex), around the clause (those fac-
tors ensuring the text cohesion and coherence from the textual perspective), and
beyond the clause (grammatical metaphor). In this chapter, we will discuss the
SFL approach to absolute clauses. First, we will offer a sketch of functional syn-
tactic theory, then define absolute clauses following the idea of cline in the frame-
work of SFL, and finally suggest the identification criteria of absolute clauses.

© Springer-Verlag Berlin Heidelberg 2015 31


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_3
32 3  Approaching Absolute Clauses …

3.2 Functional Structure of Clause

Like traditional syntax, functional syntax consists of two parts: form and func-
tion. Traditional syntactic analysis is a maximal bracketing approach, i.e., imme-
diate constituent analysis (IC analysis), concerning about what meanings a form
can express; functional syntactic analysis is a minimal bracketing approach, i.e.,
ranked constituent analysis, concerning about what forms can realize a certain
meaning. Traditional grammar constitutes a set of syntactic structure, while func-
tional grammar, three sets of syntactic structure. For example, the syntactic struc-
tures of example 3-1 are shown in Fig. 3.1:
3-1 Shortages made life difficult, to the point of exasperation. (BNC_NA)
Traditional grammar includes absolute clauses into the category of non-finite
and verbless adverbial clauses with an explicit subject. They have their own sub-
ject–predicate structures. According to SFL, absolute clauses are at the clause rank
in rank scale. Being clauses, they can be analyzed in transitivity structure, mood
structure, thematic structure, and information structure.

3.2.1 Transitivity Structure

“The clause—like any other grammatical unit—is a multifunctional construct con-


sisting of three metafunctional lines of meaning” (Halliday and Matthiessen 2004:
58, 168). Ideational metafunction is realized from human’s construing the external
and internal experience through meaning (Halliday and Matthiessen 1999). The
transitivity structure realizing ideational function typically includes participants,

shortages made life difficult

Traditional Syntax Subject Predicate Object Complement

Transitivity Goal
Actor Process: Material
Structure Carrier Attribute

Mood Subject Finite Main Verb Complement


Functional
Structure Mood Residue
Syntax
Thematic
Theme Rheme
Structure

Information Given New

Fig. 3.1  Traditional and functional structures


3.2  Functional Structure of Clause 33

processes, and circumstances. The process is the core element, realized by verbal
groups, and participants are close to this core element, realized by nominal groups.
Therefore,
We can say that the configuration of process + participants constitutes the experiential
centre of the clause. Circumstantial elements augment this centre in some way — tempo-
rally, spatially, causally and so on; but their status in the configuration is more peripheral
and unlike participants they are not directly involved in the process. Circumstances are
realized by adverbial phrases or prepositional phrases. (Halliday and Matthiessen 2004:
176)

SFL distinguishes six types of process. Non-finite verbs realizing the six types
of process can all constitute absolute clauses. In the absolute clauses constituted
by verbless clauses, the relational verbs are omitted, with minimal groups, adver-
bial groups, or prepositional phrases realizing attribute. For example,
3-2a. Hands shaking, I walked deeper into the house, Benjamin behind me.
(BNC_FIC) (Material)
b. The Lord being good, Joseph hopes for a new son for the House of David.
(COCA_FIC) (Relational)
c. It’s always been security problems, people talking about war. (COCA_
NEWS) (Verbal)
d. Tiara sat, people watching, as she sipped her punch. (COCA_FIC)
(Behavioural)
e. Jo Ellen could hardly stand it, people knowing her name, asking her for rec-
ipes, hugging her. (COCA_MAG) (Mental)
f. I went upstairs to my room, there being no other place. (COCA_FIC)
(Existential) (Fig. 3.2)

3.2.2 Mood Structure

Interpersonal function is the function to enact the social process or human rela-
tionship using language. “The principal grammatical system is that of MOOD.”
(Halliday and Matthiessen 2004: 106) The mood structure consists of the mood
and the Residue. Mood is the constituent formed by Subject and Finite, and the
main verb falls in the Residue. The mood structure of absolute clauses comprises
the Subject and the Residue, there being no Finite. There is no mood in a clause
without Subject and Finite, the Finite playing the decisive part. This is to say that
absolute clauses are clauses without mood. For example,
3-3 Time permitting, we shall discuss the matter. (COCA_FIC)
The absolute clause Time permitting in 3-3 contains a Subject and a main verb,
but not a Finite, hence no mood. In interrogative clauses, the Finite is the element
that can be moved before the Subject. In Time permitting, it is the Finite does or is
that is omitted. The finite clause is Time is permitting or Time permits (Fig. 3.3).
34 3  Approaching Absolute Clauses …

hands Shaking,
I walked deeper into the house, Benjamin behind me.
Medium Process: Material

The Lord being good,


Joseph hopes for a new son for the House of David
Carrier Process: Relational Attribute

people talking about war


It’s always been security problems,
Sayer Process: Verbal Verbiage

people watching,

Tiara sat, Process: as she sipped her punch.


Behaviour
Behavioural

People knowing her name, asking her for recipes,


Jo Ellen could hardly stand it,
Senser Process: Mental Phenomenon hugging her.

there being no other place.


I went upstairs to my room,
Process: Existential Existent

Fig. 3.2  Transitivity structure of absolute clauses

Time permitting, we shall discuss the matter.

Subject Main verb

Residue

Fig. 3.3  Mood structure of absolute clauses

In the absolute clauses constituted by verbless clauses, the complement func-


tions as the residue, the finite form being attributive relational clauses. For
example,
3-4a. A man stumbled out of the alley, gun in hand. (COCA_FIC)
b. The baby waves, her hand turned backwards, knuckle-side out. (COCA_FIC)
c. The interview over, the detective followed Yaz home. (COCA_SPOK)
In 3-4, the complements of the three absolute clauses are played by preposi-
tional phrase, adverb, and adjective, respectively. Grammarians such as Fowler
(1965: 5), Onions (1971: 76), and Kane (1983: 755) refer to clauses of this type
of structure as reduced absolute clauses, with participial elements omitted, mostly
being or having been.
3.2  Functional Structure of Clause 35

Mayhew having concluded his narration, Ahab put such questions to him.

Theme Rheme Theme Rheme

Given New Given New

Theme Rheme

Given New

Fig. 3.4  Thematic and information structures of absolute clauses

3.2.3 Thematic and Information Structures

The textual metafunction “is a second-order mode of meaning” (Halliday and


Matthiessen 1999: 398). It enables the realization of the ideational and interper-
sonal metafunctions, offering environment for the construing of experience and
the enacting of social process, and is realized through the thematic structure and
information structure. The former consists of the Theme and the Rheme, the two
constituting a message. The latter is composed by the Given and the New, the two
constituting an information unit. Absolute clauses have also their own thematic
and information structures. In the thematic structure, the nominal group func-
tioning as the subject is the Theme, and the non-finite clause or verbless clause
functioning as the Residue is the Rheme. In the information structure, the nominal
group functions as the unmarked Given information, and the non-finite clause or
verbless clause functions as the unmarked New information. For example,
3-5 
Mayhew having concluded his narration, Ahab put such questions to him.
(COCA_FIC) (Fig. 3.4)

3.3 Functional Structure of Clause Complex

The internal structure of a clause is a multifunctional construct. Two clauses can


be connected through some certain logico-semantic relations to form a clause
complex, one of which is the primary clause and the other, secondary. The rela-
tionship between the two clauses constituting a clause complex is determined by
two types of relationship: interdependency, which is further identified as parataxis
and hypotaxis, and logico-semantic relation, which is largely specified as expan-
sion and projection. In parataxis, the two clauses are equal in status. Both clauses
tend to stand independently, to construe a proposition, which can have a tag ques-
tion and different mood choices. However, they are not isolated from each other
because the relationship between them is structural. The first clause in a paratactic
36 3  Approaching Absolute Clauses …

clause complex is the initiating clause, which adopts the tone 3, indicating that
there is still other information following, and the second clause is the continuing
clause, which adopts the tone 1, indicating that the sentence ends here. However,
two independent clauses are cohesive, both adopting tone 1. In hypotaxis, the two
clauses are not equal in status. Only one of the two clauses, the dominant clause,
can be independent from the other, and the dependent clause functions as the
modification of the primary clause. The dominant clause construes a proposition
and can have a tag question, but the dependent clause cannot. A hypotactic clause
complex has no corresponding cohesive equivalents. However, two independent
clauses can realize the logico-semantic relationships of time, cause, condition,
and concession with no conjunctions realizing the relator, but sometimes there are
conjunctive adverbs to connect the two clauses. Because of the unequal status, the
position of the dependent clause depends on the requirement of the textual con-
struction; it can be positioned before or after the dominant clause. The dependent
clause positioned before the dominant clause is represented as the Theme of the
clause complex, which is made prominent through topicalization, and adopts the
tone 4 or tone 1, while the dependent clause positioned after the dominant clause
has no topic status and adopts the tone 1.
In expansion, the secondary clause elaborates, extends, or enhances the pri-
mary clause, and in projection, the primary clause projects a locution or an idea.
“Expansion relates phenomena as being of the same order of experience, while
projection relates phenomena to phenomena of a higher order of experience
(semantic phenomena—what people say and think).” (Halliday and Matthiessen
2004: 377) Interdependency and logico-semantic relations intersect to form a rela-
tion network of clause complexes, as shown in Fig. 3.5.
According to the interdependency and logico-semantic relation, Halliday
(1994: 220) distinguishes ten basic relationships of absolute clauses. See Table 3.1
In hypotaxis, “the dominant element is free, but the dependent element is not”
(Halliday 1994: 221). The semantic types of the hypotactic clause in the clause

Fig. 3.5  The system network hypotaxis


of clause complexes TAXIS

parataxis idea

clause projection
LOGICO- locution
elaborating
SEMANTIC TYPE
extending
expansion
stop enhancing
RECURSION

go on
3.3  Functional Structure of Clause Complex 37

Table 3.1  Basic types of clause complex (Halliday 1994: 220)


(i) Paratactic (ii) Hypotactic
(1) (a) John didn’t wait; John ran away, which surprised
Expansion Elaboration he ran away. everyone.
1 = 2 α = β
(b) John ran away, John ran away, whereas Fred
Extension and Fred stayed behind. stayed behind.
1 + 2 α + β
(c) John was scared, John ran away, because he was
Enhancement so he ran away. scared.
1 × 2 α × β
(2) Projection (a) John said, “I’m John said he was running away.
Locution running away.” α “β
1 “2
(b) John thought to himself, John thought he would run away.
idea “I will run away.” α ‘β
1 ‘2

complex are the circumstances in the transitivity system of clause. “Circumstances


augment the configuration of process + participants in the clause in terms of either
projection or expansion.” (Halliday and Matthiessen 2004: 367).
3-6a. She loved her daughter with all her heart. (COHA_FIC)
b. Water had to be ferried from the nearest hydrant using a bowser.
(BNC_NEWS)
The prepositional phrase with all her heart in 3-6a augments the circumstance of
the clause, and the non-finite clause using a bowser in 3-6b expands the clause,
opening up a clause complex. 3-6 shows that the circumstance of a clause contains
only one minor process, hence not being able to realize a figure or enact a proposi-
tion or proposal, nor being able to represent an information unit. On the contrary,
a clause complex can always be assigned a clause status to the expansion and pro-
jection elements. The augmented element has the potential to realize a figure in
experiential, interpersonal, and textual metafunctions. Circumstances are a part of
a clause, and clauses in clause complexes are a part of the serial structure. In cre-
ating a text, “we choose between augmenting a clause ‘internally’ by means of a
circumstantial element and augmenting it ‘externally’ by means of another clause
in complex” (Halliday and Matthiessen 2004: 369).
Clause complexes are called compound sentences in traditional grammar,
which adopts a bottom-up approach, starting from language form to study the syn-
tax and the meaning of compound sentences. SFL adopts a top-down approach,
starting from the meaning to study the logico-semantic relations of clause com-
plexes. If the logico-semantic relation is clear between the two clauses in a clause
complex, an explicit conjunctive expression is not necessary at the lexico-gram-
matical level. “Semantically, the effect of combining clauses into a clause complex
is one of tighter integration in meaning” (ibid.: 365). A sequence at the semantic
38 3  Approaching Absolute Clauses …

level can be realized at the lexico-grammatical level by two independent clauses


cohesively linked together, by a paratactic or hypotactic clause complex composed
of two finite clauses, by a hypotactic clause complex composed of a finite clause
and a non-finite clause, or even by a single clause through nominalization. In this
process, the degree of meaning integration increases.
The sequence of projection or expansion at the semantic level can be realized
by two clauses structurally combined together to form a clause complex, or by
two independent clauses cohesively linked together, or by a clause and a group or
phrase functioning as a circumstance of this clause. These different lexico-gram-
matical forms constitute a cline, with one simple clause at one end, and two inde-
pendent clauses at the other, and the clause complex in between, but not at one
point, rather including two areas. One area is close to the circumstance, where one
clause of the two depends on the other, hence unequal in status, realizing hypo-
taxis, and the other area is close to the cohesive sequence, where the two clauses
are interdependent, hence equal in status, realizing parataxis.
Absolute clauses belong to the category of non-finite clauses. Following the
SFL theory of clause complex, non-finite clauses can only realize hypotaxis. The
three function types of absolute clauses, i.e., explanation, attendant circumstance,
and clausal adjuncts we discussed in Chap. 2 correspond, respectively, to the non-
finite clauses of the expansion types, i.e., elaboration, extension, and enhancement
in the SFL clause complex. The SFL theory offers a reasonable grammatical basis
for the explanation type of absolute clauses in traditional grammar. For example,
3-7 Guests in dinner clothes stood on the brink, the men laughing, the ladies
giggling. (COHA_FIC) (elaboration)
3-8 Butterflies as big as birds fell down from the branches, wings flopping
limply. (BNC_FIC) (extension)
3-9a. Job done, the bat fluttered back to its perch. (COCA_FIC) (enhancement:
time)
b. Compromise having failed, there was left only force. (BNC_ACAD)
(enhancement: cause)
c. Injuries permitting, he should finish this season approaching 600 career
games. (BNC_MISC) (enhancement: condition)

3.4 Rank Status of Absolute Clauses

Although non-finite clauses are at the same rank as clauses, they have some certain
specific features different from finite clauses in syntactic function. For example,
the -ing forms of verb include present participle and gerund, and present partici-
ples share more features of clauses, while gerunds share more features of nominal
groups. In this section, we will discuss the rank status of absolute clauses accord-
ing to the theory of cline of SFL.
3.4  Rank Status of Absolute Clauses 39

3.4.1 Cline

Cline refers to those language units which cannot be included in any language
category (Halliday and Matthiessen 1999). “To categorize linguistically is to put
together the linguistic symbols and the human experiences these symbols rep-
resent so as to set up categories.” (Yang 2007: 50) The concept of “category”
originates from the Aristotelian Classical Philosophy, according to which cat-
egory is defined in terms of sufficient and necessary features, and all members
in a category are equal in status. Influenced by Wittgenstein’s thinking of family
resemblance, cognitive linguistics has raised the prototype theory of semantic cat-
egorization, pointing out that there are degrees of membership based on degrees of
similarity. Unlike the classical theory and the cognitive theory, the theory of SFL
is not oriented toward philosophy or logic, but toward language and language use,
in other words, concerning “with how meaning is construed in naturally occurring
text” (Halliday and Matthiessen 1999: 72).
According to SFL, there are no neat boundaries between categories, and
the members of the two categories form a cline. For example, “the distinction
between closed system patterns and open set patterns in language is in fact a
cline” (Halliday 1961[2002]), and that between participants and circumstances is
also a cline (Halliday and Matthiessen 1999) and the nominal groups in preposi-
tional phrases are indirect participants (Halliday 1994: 150). The notion cline is a
major indicator to distinguish SFL from formal grammars, and cline can be seen
in many concepts of SFL. For example, “material, mental, and relational processes
are the main types of process in the English transitivity system” (Halliday and
Matthiessen 2004: 171), and on the borderlines between them are the behavioral,
verbal, and existential processes. These six processes form a cline.
The idea of cline is helpful to explain many linguistic phenomena that formal
grammars have difficulty explaining. For example, influenced by Aristotle’s cat-
egory theory, people distinguished various language units and developed many
rules to categorize language. Take adverbial participles as an example. According to
­traditional grammar, the subject of adverbial participles should be the subject of the
main clause, as in 3-10a; if not, it should have its own explicit subject, hence abso-
lute clauses, as in 3-10b. However, there are also some adverbial participles with the
subject not being that of the main clauses, hence dangling participles, as in 3-10c.
3-10a. Walking in the streetbeside Marla, Toni felt something again. (COCA_FIC)
b. Weather permitting, the bar is open from 6 p.m. to midnight. (COCA_NEWS)
c. Judging from Roy’s earnest manner, this was serious stuff .(COCA_FIC)
Traditional grammar refers to the participial phrases in 3-10a and 3-10c as free
adjuncts, the former being related free adjunct and the latter, unrelated free adjunct
(dangling participle). Aristotle’s category theory cannot distinguish judging from
walking, and the prototype theory of cognitive science considers judging as the
non-core member of a category and therefore has to give a semantic explanation
to this phenomenon. According to formal grammar, judging has no logical subject
because it has become a fixed phrase.
40 3  Approaching Absolute Clauses …

The notion of cline of the SFL category theory can offer a reasonable explana-
tion to judging in 3-10c. Participial phrases constitute non-finite clauses, realiz-
ing circumstances. However, the core members of the category of circumstances
are prepositional phrases. It is a cline between non-finite clauses and preposi-
tional phrases. Phrases can fall into the group rank or the word rank (Yang 2001).
Generally, verbs are in requirement of a subject more than prepositions and tend
to be grammaticalized into prepositions. Grammaticalization “is confined to the
development from lexical to grammatical forms” (Heine and Kuteva 2004: 4). For
example, considering, regarding, and concerning in English have all grammatical-
ized from verbs to prepositions. The English preposition during originates from
the old French verb durer (meaning to continue), which was loaned into English
in the Middle Ages, the present participle form being duren. Now, the word has no
traces of a verb and has completely grammaticalized into a preposition. Therefore,
it is a cline from participles to prepositions. For example,
3-11a. Looking at the picture, I could feel his love from far away. (COCA_FIC)
b. Considering the times, it seemed destined for success. (COCA_NEWS)
c. During the war, everything here went to pieces. (COCA_FIC)
The participle looking in 3-11a is the strongest in motion, and looking at the
picture is a non-finite clause. It is in strong requirement of a subject. In 3-11c, dur-
ing is the weakest in motion, and During the war is a prepositional phrase. It is not
in strong requirement of a subject. In 3-11b, considering lies in between looking
and during in motion. It is an internal mental activity and has no obvious external
relationship with the subject of the main clause.
There are also many clines in the logico-semantic relations of non-finite
clauses. For example,
3-12a. I worked for a local firm at that time, selling office equipment. (Halliday
and Matthiessen 2004: 404)
b. She lay awake for some time, puzzling over Fen’s behaviour. (BNC_FIC)
c. I’ve seen eight people leaving the Stenness Hotel. (BNC_SPOK)
3-12a is a clause complex, realizing elaboration. 3-12c is a simple clause with an
embedded non-finite clause, realizing modification. 3-12b lies in between elabora-
tion and embedding. If it is considered as elaboration, the whole structure is a clause
complex; if it is considered as embedding, the whole structure is a simple clause.

3.4.2 Rank Status

Non-finite clauses are at the clause rank. However, they have some specific fea-
tures different from finite clauses, e.g., they do not realize parataxis in clause com-
plexes. Despite the fact that both traditional grammar and SFL ascribe absolute
clauses into the category of non-finite clauses, they are not mere non-finite clauses
with a subject. They are different from non-finite clauses in that they do not need
3.4  Rank Status of Absolute Clauses 41

conjunctive expressions to realize the relator, and they are not co-referential with
the subject of the main clauses. In this section, we will discuss the rank status of
absolute clauses in terms of the notion of cline.
Halliday (1985: 193, 1994: 216) classifies lexico-grammatical units into
five basic constituents in rank, including clause complex, clause, phrase/group,
word, and morpheme. However, he later excludes clause complex from the rank
scale, leaving only four constituents: clause, phrase/group, word, and morpheme
(Matthiessen and Halliday 1997[2009]: 71–72). This is because the clause com-
plex is not a lexico-grammatical unit, and the four grammatical units can all form
complexes (Yang 2003: 44; Halliday and Matthiessen 2004: 9). There are not dis-
tinctive boundaries between the four ranks; rather, they form a cline. For example,
at the group/phrase rank, groups are bloated words and are closer to words in rank,
and phrases are compacted clauses and are closer to clauses in rank. Although
non-finite clauses are at the clause rank, their core element, non-finite verbs, deter-
mines that they are in between finite clause and group in the cline of rank scale.
Non-finite verbs distinguished by traditional grammar include present parti-
ciple, past participle, infinitive, and gerund. SFL does not make such a distinc-
tion, but points out in discussing hypotaxis that “the dependent clause often has an
explicit Subject of its own; when this can show a contrast in case, it appears either
in oblique (e.g. him) or in possessive (e.g. his) form” (Halliday and Matthiessen
2004: 421). For example,
3-13 With him/his taking time off, everyone has to work harder. (ibid)
According to the traditional grammar, in such a structure with him taking time
off, him is the object of the preposition with, and taking time off is the comple-
ment of him, forming a logical subject–predicate structure, taking being a present
participle. In with his taking time off, his taking time off is the object of with, form-
ing a prepositional phrase, taking being a gerund. “One problem of terminology
and analysis arising in connection with both free adjuncts and absolutes is that
of distinguishing present participles from gerunds.” (Stump 1985: 11) “Halliday
appears to treat the ‘conjunctive preposition’ as a conjunction for the purposes of
the clause analysis.” (Butler 2003: 269, 271) In with him taking time off, him tak-
ing time off is a non-finite clause, and with is a conjunctive preposition, realizing
relator. In with his taking time off, the possessive pronoun his “reflects the earlier
status of these non-finite clauses as rank-shifted” (Halliday and Matthiessen 2004:
421). In his taking, taking is the nominalization of take, closer to group/phrase in
rank. Therefore, his taking is a constituent of the finite clause, and it should not be
considered as a hypotactic clause in a clause complex. However, it still construes a
figure at the semantic level and construes a sequence with the finite clause. Let us
see example 3-14:
3-14a. When he arrived, they would figure out something together. (COCA_FIC)
b. The proper season arriving, the vine is again pruned, and again eight or
ten times as many buds are retained as the plant can nourish. (COHA_NF)
c. When arriving at the station he asked for two tickets.(COHA_MAG)
42 3  Approaching Absolute Clauses …

d. Arriving at a suitable pitching site, I would erect the tent and Pete would
fetch water and have a brew ready by the time we piled in. (BNC_MISC)
e. With winter arriving, they turned up one chilly night at 17th Street and
Columbia Road NW… (COCA_NEWS)
f. On arriving at the lounge, he saw Brother Emil Sanger already seated at
the far end of the circle of chairs. (COCA_FIC)
g. Upon my arriving there the Gypsies swarmed out from their tents.
(COHA_NF)
h. On arrival at Honey Cottage, Yanto introduced the two girls. (BNC_FIC)
i. On his arrival at Leipzig in the autumn of 1865, … he was in low spirits.
(BNC_ACAD)
The hypotactic clause in 3-14a is finite, conjunction when realizing the relator.
The hypotactic clause in 3-14b is an absolute clause, without a conjunctive expres-
sion realizing the relator and the logico-semantic relation being explicit and arriv-
ing is a present participle, the logical subject being a zero case noun. Even if there
is a conjunctive expression, as in 3-14e, the non-finite clause with subject func-
tions as the complement of with, and arriving is still a present participle. In 3-14c,
the hypotactic clause is non-finite, and the use of conjunction when indicates that
arriving is a present participle. In 3-14d, there is no conjunctive expression real-
izing the relator, and it may be a conjunction, as in 3-14c, or a preposition, as in
3-14f. The preposition on in 3-14f realizes relator, and the non-finite verb arriving
functions as the complement of the preposition on and this means that arriving has
begun to be nominalized. The arriving in 3-14g is modified by the possessive pro-
noun my and is further nominalized. The arrival in 3-14h and 3-14i is the nomi-
nalization of verb, constituting a preposition phrase with on. Although arrival is a
nominal group, it still construes a figure at the semantic level.
The verbal group arrive in 3-14 is gradually nominalized from 3-14a to 3-14i,
and in the process of nominalization, the degree of grammatical metaphor is
increasing. Therefore, from the point of rank scale, the nominalization process
from 3-14a to 3-14i is a downward rankshift from clause to group/phrase. Verbs
are the core elements of clauses, and they determine the rank status of clauses;
nouns are the core elements of nominal groups, and they determine the rank status
of the nominal groups. The cline of rank from verbal group to nominal group is
shown in Fig. 3.6.
This means that absolute clauses should be positioned between finite clauses
and non-finite clauses introduced by a conjunction along the cline of rank. This
analysis is also true for verbless clauses.

Fig. 3.6  Cline from clause clause finite verb


to group
participle

gerund

nominal group noun


3.5  Formation Requirement and Identification Criteria of Absolute Clauses 43

3.5 Formation Requirement and Identification


Criteria of Absolute Clauses

The purpose of this section is to probe into the requirement for forming absolute
clauses and then to find out the criteria for identifying absolute clauses from the
SFL perspective.

3.5.1 Formation Requirement

Traditional grammar defines absolute clauses as adverbial clauses with an explicit


subject, in terms of which the formation of absolute clauses should fulfill the fol-
lowing three requirements: (1) functioning as adverbial adjuncts of the main
clauses, (2) having an explicit subject, and (3) having no explicit conjunctive
expressions. However, according to the analysis in Chap. 2, absolute clauses of
explanation are not adverbial adjuncts, but appositives. Therefore, functioning as
adverbial adjuncts is not a necessary requirement for forming absolute clauses.
Traditional grammarians (e.g., Visser 1972: 1147) generally hold that the subject
of absolute clauses is not co-referential with that of the main clauses. In other
words, the implicit subject of non-finite clauses functioning as adverbial adjunct
is co-referential with that of the main clauses. If they are not co-referential, the
subject of non-finite clauses should be explicit, forming absolute clauses. It is
sure that sometimes the subject of an absolute clause and that of the main clause
are co-referential. Kortmann (1991: 92) distinguishes four possible types of sub-
jects of absolute clauses: completely non-co-referential subjects, subjects display-
ing constituent, part-whole, full co-reference. The subject of most of the absolute
clauses is of part-whole or full co-referential relation with that of the main clauses.
Therefore, non-coreference is still not a necessary requirement for forming abso-
lute clauses. For example,
3-15a. It was his sense of history, part romantic, part Christian. (BNC_ACAD)
(part-whole)
b. Rob slammed the door when he came in, the bastard doing everything in
his power to wake us up. (Berent 1975: 14) (full coreference)
c. Their patron, St Anthony, was the Egyptian hermit, he having been held
to foster the growth of herbs in the desert. (Scheurweghs 1969: 164) (full
coreference)
This shows that in the above three requirements, only “having no explicit con-
junctive expressions” makes the sufficient and necessary requirement. Despite the
fact that the subjects of absolute clauses in 3-15 are completely or partially co-
referential with those of the main clauses, the co-referential subjects are not direct
participants of the main verbs in the main clauses. These co-referential subjects
refer back to the subjects of the main clauses. Reference is a grammatical cohesive
44 3  Approaching Absolute Clauses …

device, and it functions to make the two clauses cohesively linked together at the
textual level.
In addition, the identification of absolute clauses in the traditional sense is
writing based. An absolute clause is always separated from the other part of the
sentence by a punctuation mark. However, the conjunctive adverb however in the
above example 2-1a will still be an absolute constituent even if it is not separated
by a comma. The reason is that whether there is a comma or not, an absolute con-
stituent is not an immediate participant of the relevant clause, and the comma is
only an explicit indicator. For example, the comma in example 2-1b cannot be
left implicit because if there is no comma, the adjective alone may be mistaken to
function as the modifier of Sara although this is ungrammatical since the adjective
alone itself cannot assume the role of modifier. On the other hand, the adverb is
flexible in a clause, and it is always separated from other constituents by a comma.
Whether it is separated or not, the syntactic function of the adverb will not change.
Therefore, it cannot be taken as the identification principle or the formation
requirement of absolute clauses to say that absolute clauses are “not grammati-
cally tied to the sentence” (Kane 1983: 754) or “not explicitly bound to the matrix
clause syntactically” (Quirk et al. 1985: 1120). Absolute clauses can be defined in
the framework of SFL as follows:
Non-finite clauses consisting of a nominal group that is not a direct participant in the primary
clauses and a nonfinite phrase not introduced by an explicit conjunctive expression.

Non-finite clauses can realize various syntactic functions. Since absolute


clauses are not confined to some certain syntactic functions, they have also the
potential to realize various syntactic functions. For example, the subject in both
People smoking in public is prohibited and People smoking in public will be pun-
ished is constituted by a nominal group and a non-finite element. However, the
nominal group people and the non-finite element smoking in public constitute a
subject–predicate construction embedded in the finite clause to function as the
subject in the former, while the nominal group people and the non-finite element
smoking in public do not constitute a subject–predicate construction in the latter
where smoking in public is embedded in the core noun people to function as its
post-modifier. The same construction People smoking in public in People smoking
in public, the alarm will be switched on automatically functions as a circumstance.
In the three sentences, only the nominal group people in the second sentence real-
izes a participant in the primary clause, hence not in agreement with the defini-
tion of absolute clauses. The nominal group people in both the first and the third
sentences functions as the subject of non-finite verbs and has nothing to do with
the main verbs in the primary clauses, hence in agreement with the definition of
absolute clauses. Since absolute clauses can function as appositives, they can also
function as subjects. In the next section, we will investigate the criteria for identi-
fying absolute clauses.
3.5  Formation Requirement and Identification Criteria of Absolute Clauses 45

3.5.2 Identification Criteria

At the lexico-grammatical level, absolute clauses are constituted by nominal


groups and non-finite phrases. However, not all constructions consisting of a nom-
inal group and a non-finite phrase are absolute clauses. In a clause complex, the
relation between clauses is “a relation between processes” (Halliday 1994: 216),
and “the study of clause complexes, especially those which contain non-finite
clauses, is better to focus on ‘the relation between processes’” (Yang 2003: 50).
The logico-semantic relations between the two finite clauses in a clause complex
are usually realized by conjunctions, and “in clause complexes involving non-
finite clauses, the non-finite element can be considered as one that functions as a
conflation of process and relator” (ibid.: 57). This is also true for absolute clauses
as non-finite clauses. Therefore, the criteria for identifying absolute clauses can be
worked out from the three functional elements: relator, participant, and process.

3.5.2.1 Relator

Relator is realized by conjunctions or conjunction groups (Halliday and


Matthiessen 1999: 177). In clause complexes involving non-finite clauses, rela-
tor can be realized by conjunctions or conjunction groups, prepositions or prepo-
sitional groups, and zero conjunctive expressions. In clause complexes involving
absolute clauses, relator is realized by zero conjunctive expressions; that is, there
are no explicit conjunctive expressions. Let us see examples 3-16–3-20:
3-16a. Pastor Kramer steps out of the car, hat in hand. (COCA_FIC)
b. Time permitting, we shall discuss the matter. (COCA_FIC)
3-17a. He would not insist on having two ballot boxes, one for yeses, one for
noes. (BNC_MAG)
b. She could hear voices ahead, a girl squealing, a man laughing. (BNC_FIC)
3-18a. He heard Craig chuckling inside his helmet. (COCA_FIC)
b. They consider Christmas to be a pagan holiday. (COCA_SPOK)
3-19a. Tom flying planes badly can be dangerous. (Yang, 2003: 111)
b. People living nearby were told to stay indoors. (BNC_NEWS)
3-20 One was knitting, the other sucking a mint. (COCA_FIC)
3-16a and 3-16b are typical absolute clauses in traditional grammar. The former
is an attendant circumstance and the latter, an adverbial clause, realizing the rela-
tionship of cause or condition. From the perspective of SFL, 3-16a realizes hypo-
tactic extension and 3-16b, hypotactic enhancement. Both absolute clauses have
no explicit conjunctive expressions, and the difference lies in that the finite form
of the former is a paratactic clause complex of extension linked by and and that of
46 3  Approaching Absolute Clauses …

the latter, a hypotactic clause complex introduced by a hypotactic conjunction. For


example,
3-16a. Pastor Kramer steps out of the car, and a hat is in his hand.
b. If time permits, we shall discuss the matter.
3-17 does not differ from 3-16 in form. Although traditional grammar includes
the non-finite clause in 3-17 as an absolute clause, it does not function as an adver-
bial adjunct; rather, it is a nonrestrictive relative clause, functioning as expla-
nation. Both relative clauses and adverbial clauses belong to the category of
subordinate clauses, but only adverbial clauses “can be positioned initially, medi-
ally, and finally” (Quirk et al. 1985: 1124), relative clauses can only be positioned
after the antecedent. The non-finite form of the finite clause in 3-17a is a nonre-
strictive relative clause, with the subject one of the absolute clause functioning as
the subject of the nonrestrictive relative clause, and the relative pronoun which as a
post-modifier. The finite form of the absolute clauses in 3-17b is a girl was squeal-
ing or (a girl squeals), a man was laughing (or a man laughs). In fact, a girl
squealing, a man laughing is the appositive of voices syntactically, functioning as
explanation or exposition of the antecedent voices. This sentence can be rewritten
as She could hear a girl squealing and a man laughing. According to SFL, the two
non-finite clauses can also be changed into nonrestrictive relative clauses, realizing
the relation of hypotactic elaboration, the relative pronoun which functioning as
the subject of the nonrestrictive relative clause, and a girl squealing, and a man
laughing as the complement.
3-17a. He would not insist on having two ballot boxes, one of which is for yeses,
and the other is for noes.
b. She could hear voices ahead, which are a girl squealing, and a man
laughing.
The non-finite clauses in 3-18 are projected non-finite clauses, and grammati-
cally, they function as the complement of the main verbs. Both the two sentences
in 3-19 contain the construction of “nominal group + non-finite element,” but
this construction in 3-19a functions as the subject of the main verb can be. In the
“nominal group + non-finite element” construction of 3-19b, the nominal group
functions as the subject of the main verb were told, and the non-finite element
functions as the post-modifier of the nominal group. The hypotactic clause in a
clause complex of expansion can be considered as the circumstance of the primary
clause in transitivity and that in a clause complex of projection, as a participant
of the main verb. Similarly, the embedded clause functioning as subject can also
be seen as a participant of the main verb. “The logico-semantic relation between
clauses in clause complexes is mainly determined by conjunctions” (Yang 2003:
75). Finite clauses realizing circumstances require conjunctions, and finite clauses
realizing participants also have conjunctions realizing relator, but this kind of con-
junctive expressions can always be omitted. There are two reasons. One is that
there is some certain logico-semantic relation between the two clauses in a clause
complex, and the explicit conjunction can not only indicate the hypotactic relation
3.5  Formation Requirement and Identification Criteria of Absolute Clauses 47

between the two clauses but also make the logico-semantic relation explicit. The
other is that the logico-semantic relation between the secondary clause functioning
as participant and the primary clause is unique, and it does not need to be real-
ized by explicit conjunctions, the clause functioning as complement being a pro-
jected one. The classification of projection is not realized by conjunctions but by
projecting verbs. Clauses functioning as subject are embedded clauses, which can
be introduced by the conjunction that or not. In general, finite clauses realizing
circumstances or participants are linked by conjunctions. When the finite clauses
realizing circumstances are reduced to non-finite clauses, the conjunctions can be
left implicit on condition that the logico-semantic relation is clear. The only inter-
pretation for the explicit conjunctions is that they can make the logico-semantic
relation clear. The logico-semantic relation between a non-finite clause realizing
participants and the primary clause cannot cause misunderstanding; hence, an
explicit conjunction is no longer required. Non-finite clauses in 3-18a, 3-18b, and
3-19a can all have their finite equivalents. For example,
3-18a. He heard (that) Craig was chuckling inside his helmet.
b. They consider (that) Christmas is a pagan holiday.
3-19a. (That) Tom flying planes badly can be dangerous.
The non-finite clauses in 3-19b and 3-17 also have their finite equivalents,
introduced by relative pronoun which or who. The relative clause in the former is
nonrestrictive and that in the latter, restrictive. For example,
3-19b. People who live nearby were told to stay indoors.
According to SFL, the nonrestrictive relative clause rankshifts to function as
the modifier of the nominal group, the whole construction being a simple clause,
and that in 3-17 is a hypotactic clause, elaborating the primary clause, the whole
construction being a clause complex. Therefore, having finite equivalents can be
taken as a criterion for identifying absolute clauses. So we have the first criterion
as follows:
Criterion One: Absolute clauses have their finite equivalents introduced by
conjunctions.

3.5.2.2 Participant

The subject of an absolute clause of extension and enhancement is co-referential


with that of the primary clause, and the subject of an absolute clause of elabo-
ration is complete co-referential or partial co-referential. According to the previ-
ous discussion, non-co-reference is not the sufficient and necessary requirement
for forming absolute clauses. Whether or not being co-referential with that of the
primary clause, the subject of absolute clauses does not directly assume the func-
tion of participants syntactically. In 3-16a, the subject hat of the absolute clause of
extension is not a participant of the verbal group steps out of the primary clause.
In 3-16b, the subject time of the absolute clause of elaboration is not a participant
48 3  Approaching Absolute Clauses …

of the verbal group shall discuss of the primary clause. The subject one of the
absolute clause of elaboration is substitute, realizing a part-whole relation with the
antecedent boxes. In 3-17b, the subject a girl and a man of the absolute clauses
of elaboration can be seen as the complement of the verbal group of the primary
clause semantically, but they both are not the direct participants of the verbal
group of the primary clause.
The participant oriented criterion for identifying absolute clauses can be helpful
for better explaining the absolute clauses in 3-18 and 3-19. “When a noun phrase
intervenes between the host verb and the to-infinitive, it is often unclear whether
the phrase belongs to the host clause or the complement clause. In either case, if it
is a pronoun it is in the objective case.” (Greenbaum 1996: 350) 3-18a is a clause
complex of projection, in which Craig chuckling inside his helmet is a non-finite
clause with subject, the whole construction functioning as the complement of the
main verb heard. In this construction, Craig functions both the complement of
the main verb and the subject of the non-finite verbal group. Therefore, the whole
sentence is a combination of two clauses. From this sense, the subject Craig is
not independent from the main verb. However, semantically, what He heard is
the sound of laughing of Craig, rather than Craig himself. The subject Craig of
the non-finite clause cannot be separated from the non-finite verbal group; they
together realize the phenomenon of the mental verb heard. Jespersen (1933) refers
to the structure that can form a complete piece of communication as independent
nexus, which “forms only a part of a sentence, and thus may be either a primary in
a sentence (subject or object), a secondary (an adjunct) to a primary in a sentence,
or a tertiary in a sentence” Jespersen (1933: 309). Jespersen (1933: 310) explains
the simple nexus functioning as the complement of the main verb in 3-18 more
clearly through 3-21.
3-21 I found her gone. (Jespersen, 1933: 310)
It is clear that in 3-21 what I found is her gone. I did not find her because she
was gone.
3-18b and 3-18a are the same in form, but their syntactic structures are obvi-
ously different. In Halliday’s thinking, the construction considers Christmas to be
is a verbal group realizing the relationship of cause, hence not a clause complex.
However, Yang (2003) holds that both 3-18a and 3-18b are clause complexes of
projection. Both Craig chuckling inside his helmet in 3-18a and Christmas to be a
pagan holiday in 3-18b are projected non-finite clauses, the former being a macro-
phenomenon, construing an act and the latter, a metaphenomenon, realizing a fact.
Huddleston and Pullum (2002) try their effort to prove that in 3-18b Christmas is
the object of the verb consider, and to be a pagan holiday is the complement of
Christmas. The reason is that the whole construction will no longer be an accept-
able sentence when Christmas is removed. A complement is not an essential com-
ponent of a sentence. Since to be a pagan holiday is a complement in traditional
grammar, it is not indispensible. For example, He heard Craig in 3-18a is accept-
able, but They consider Christmas in 3-18b is not. The finite equivalent of the non-
finite clause in 3-18b is They consider (that) Christmas is a pagan holiday, from
3.5  Formation Requirement and Identification Criteria of Absolute Clauses 49

which it can be seen clearly that Christmas is a part of the hypotactic clause. In
fact, Christmas in 3-18b is not indispensible of consider, but is indispensible of to
be a pagan holiday. In addition, the nominal group Christmas following consider
cannot be omitted because the subject of the non-finite clause is not co-referen-
tial with that of the primary clause and it requires an explicit subject of its own.
However, even if the two subjects are co-referential in such a construction, the
subject of the non-finite clause is still indispensible, and then, a reflexive pronoun
like himself is required to fill the subject slot. Despite the fact that reflexive pro-
nouns have always an antecedent item, they themselves always carry an additional
meaning of emphasis, which prevents them from complete co-referential with the
antecedent. Nominative or accusative pronouns can be completely co-referential
with the antecedent, so they are acceptable grammatically and cannot form abso-
lute clauses.
The projected non-finite clause can be a fact or an act. If it is a fact, even if the
subject of the non-finite clause can be the complement of the main verb of the pro-
jecting clause, the projected clause still forms an absolute clause. This is because
a verb projecting a fact and a verb having a nominal group functioning as its com-
plement are not different in meaning. For example, the main verb believes in 3-22
means “to accept as true,” which is different from that in she believes Nathan,
meaning “trust.”
3-22 She believes Nathan to be a fine young man. (COCA_FIC)
Structurally, 3-22 is closer to 3-18a than 3-18b is. The difference between
3-18a and 3-22 lies in that in the former the secondary clause is a project act,
while in the latter, a projected fact. This can be tested by changing the non-
finite clauses into relative clauses. For example, 3-18a can be changed into He
heard Craig, who is chuckling inside his helmet, without changing the meaning.
However, if 3-22 is changed into She believes Nathan, who is a fine young man,
the meaning also changes. Therefore, projected fact clauses can all be transposi-
tioned with the primary clauses and thus can form absolute clauses.
When the projected non-finite clause construes an act, it is always not clear
whether the nominal group between the main verb and the non-finite element is
attached to the projecting clause or the projected clause. The method to test the
attribution of the nominal group is the transposition of the non-finite clause and
the primary clause, such as 3-18a. If the nominal group can be transpositioned
together with non-finite element, it can be affirmed that the nominal group is
attached to the non-finite clause, hence forming an absolute clause; if not, it is
attached to the primary clause, hence no absolute clauses formed. This can be
summarized as the participant criterion for identifying absolute clauses, that is, the
second criterion.
Criterion Two: The subject of absolute clauses is not a direct participant of
the primary clauses.
The subject in both sentences in 3-19 consists of a nominal group and a non-
finite element, but the two are different in syntactic structure. In 3-19a, Tom flying
planes badly is an embedded subject clause, and the subject Tom of the embedded
50 3  Approaching Absolute Clauses …

clause is not an participant of the main verb, rather it is independent from the pri-
mary clause, and hence forming an absolute clause. In 3-19b, People living nearby
is a nominal group, the non-finite element living nearby functioning as the post-
modifier of people which functions as the subject of the main verbal group were
told to, hence not forming an absolute clause. The absolute clause in 3-19a func-
tions as the subject of the main verbal group, so Curme (1931) refers to this kind
of absolute clauses as “absolute nominative in subject clauses.” Constructions with
an embedded absolute nominative in subject clause are not clause complexes, but
simple clauses, and therefore, transposition is out of question.

3.5.2.3 Process Oriented Criterion

Based on the above analysis, absolute clauses can be returned to their original
finite form, and the subject of absolute clauses is not that of the primary clause. In
3-20, the string the other sucking a mint is also a “noun phrase + non-finite ele-
ment” structure. It fulfills the requirement that absolute clauses have their finite
equivalents and the subject of absolute clauses is not co-referential with that of the
primary clauses. For example,
3-20 One was knitting, and the other was sucking a mint. (COCA_FIC)
Absolute clauses are clauses without Finite in form. The string the other suck-
ing a mint is the same as an absolute clause in form, but it is actually a paratactic
clause with the finite verb was omitted. A characteristic of omission is the con-
formity of syntactic functions (Zhu et al. 2001: 104). Therefore, a finite clause
with the Finite omitted does not form an absolute clause. The Finite element is
also was in the finite equivalents of the absolute clauses in 3-23. However, the was
here is not the omission of structural equatives. It is not co-referential with the ver-
bal group ’d put up in the primary clause. For example,
3-23 He’d put up a couple of mirrors, one in the downstairs hall, one upstairs on
the landing. (CLOB_L)
This indicates that the non-co-referentiality of verbs can be taken as another
criterion for identifying absolute clauses, and we have the third criterion.
Criterion Three: The verbal group of an absolute clause is not co-referen-
tial with that of the primary clause.

3.6 Summary

In this chapter, we first introduced the theory of functional syntax, including


the syntactic structure of clauses and that of clause complexes, and then gave a
SFL definition of absolute clauses. Unlike that in traditional grammar, the SFL
definition of absolute clauses does not confine the functions realized by absolute
3.6 Summary 51

clauses. Thus, absolute clauses have the same meaning potential as non-finite
clauses, for example, functioning as adjunct, appositive, subject, and comple-
ment. Following the SFL definition of absolute clauses, we developed the criteria
of identifying absolute clauses from relator, participant, and process: (1) absolute
clauses have their finite equivalents introduced by conjunctions; (2) the subject
of absolute clauses is not a direct participant of the primary clauses; and (3) the
verbal group of an absolute clause is not co-referential with that of the primary
clause. Some problems in traditional research of absolute clauses can be resolved
in terms of these three identification criteria. For example, if with in augmented
absolutes is considered as a conjunctive preposition introducing non-finite clauses,
it does not change the syntactic structure of absolute clauses nor their logico-
semantic relations with the primary clauses. However, the explicit conjunctive
expression with results in that absolute clauses are no longer independent in struc-
ture. On the other hand, if with is considered as a preposition, the complement
of the preposition is not a non-finite clause, but a nominal group containing a
post-modifier.
Due to the complexity of syntactic types, the definition and identification
­criteria developed in this chapter do not fully reflect all the characteristics of this
type of structure. For example, there are still intermediate stages between the
absolute clauses we identified and non-absolute clauses. This is a reflection of the
notion of cline in SFL. In the fifth chapter, we will discuss in detail the functional
types realized by absolute clauses in terms of the theory of functional syntax and
the independence of different functional types of absolute clauses under the notion
of cline.

References

Berent, G. B. (1975). English absolutes in functional perspective. In R. E. Grossman, L. J. San,


& T. J. Vance (Eds.), Papers from the parasession on functionalism (pp. 10–23). Chicago:
Chicago Linguistics Society.
Butler, C. S. (2003). Structure and function: A guide to three major structural-functional theo-
ries. Philadelphia & Amsterdam: John Benjamins.
Curme, G. O. (1931). A grammar of the English language: Syntax. Boston: D. C. Heath.
Eggins, S. (2004). An introduction to systemic functional linguistics (2nd ed.). London & New
York: Continuum.
Fawcett, R. P. (2000). Theory of syntax for systemic functional linguistics. Philadelphia &
Amsterdam: John Benjamins.
Fowler, H. W. (1965). A dictionary of modern English usage (2nd ed.). Oxford: Oxford
University Press.
Greenbaum, S. (1996). The Oxford English grammar. Oxford: Oxford University Press.
Halliday, M. A. K. (1961[2002]). Categories of the theory of grammar. In J. Webster (Ed.),
Collected works of M. A. K. Halliday Vol. 1: On grammar (pp. 37–94). London: Continuum.
Halliday, M. A. K. (1966). Some notes on “deep” grammar. Journal of Linguistics, 2(1), 57–67.
Halliday, M. A. K. (1985). An introduction to functional grammar. London: Edward Arnold.
Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward
Arnold.
52 3  Approaching Absolute Clauses …

Halliday, M. A. K., & Matthiessen, C. M. I. M. (1999). Construing experience through meaning:


A language-based approach to cognition. London & New York: Cassell.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2004). An introduction to functional grammar
(3rd ed.). London & New York: Edward Arnold.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2014). Halliday’s introduction to functional
grammar (4th ed.). London: Routledge.
Heine, B., & Kuteva, T. (2004). World lexicon of grammaticalization. Cambridge: Cambridge
University Press.
Huang, G.-W. (2007). Aims and principles for systemic functional syntax analysis. Foreign
Language Research, 3, 39–45.
Huddleston, R., & Pullum, G. K. (2002). The Cambridge grammar of the English language.
Cambridge: Cambridge University Press.
Jespersen, O. (1933). Essentials of English grammar. London: Allen & Unwin.
Kane, T. C. (1983). The oxford guide to writing: A rhetoric and handbook for college students.
New York: Oxford University Press.
Kortmann, B. (1991). Free adjuncts and absolutes in English: Problems of control and interpre-
tation. London: Routledge.
Matthiessen, C. M. I. M., & Halliday, M. A. K. (1997[2009]). Systemic functional grammar: A
first step into the theory. Sydney: Macquarie University Press.
Onions, C. T. (1971). Modern English syntax (7th ed.). London: Routledge & Kegan Paul.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London & New York: Longman.
Scheurweghs, G. (1969). Present-day English syntax: A survey of sentence patterns. London:
Lowe & Brydone.
Stump, G. T. (1985). The semantic variability of absolute constructions. Dordrecht: D. Reidel
Publishing Company.
Visser, F Th. (1972). Historical syntax of the English language. Leiden: E. J. Brill.
Yang, B.-J. (2001). On preposition from the functional perspective. Journal of Foreign
Languages, 1, 47–53.
Yang, B.-J. (2003). A study of non-finite clauses in English: A systemic functional approach.
Beijing: Foreign Language Teaching and Research Press.
Yang, B.-J. (2007). Realization and meaning of cline in non-finite clauses. Foreign Language
Research, 3, 50–54.
Zhu, Y.-S., Zheng, L.-X., & Miao, X.-W. (2001). A contrastive study of cohesion in English and
Chinese. Shanghai: Shanghai Foreign Language Education Press.
Chapter 4
Research Design

In the last chapter, we defined absolute clauses in the theoretical framework of


SFL and developed three criteria for identifying absolute clauses. This chapter is
the overall design of the research, including research questions, research methods,
and data collecting and processing methods, etc.

4.1 Research Questions

Absolute clauses in traditional grammar can be considered as free adjuncts with


subject. Free adjuncts are non-finite clauses functioning as adjuncts. Accordingly,
absolute clauses and free adjuncts are both included in the category of non-finite
adverbial clauses. Traditional grammar defines absolute clauses in terms of spe-
cific syntactic functions. In other words, the definition of absolute clauses in the
traditional sense is function-labeled. The research focus of SFL is language mean-
ing, and language form is the realization of meaning. Non-finite clauses have the
potential to realize multiple syntactic functions. As non-finite clauses, absolute
clauses also have the potential to realize multiple syntactic functions. At the lex-
ico-grammatical level, absolute clauses are composed of nominal groups and non-
finite elements. However, not all constructions composed of nominal groups and
non-finite clauses are absolute clauses.
Hypotactic clause complexes can be double analyzed in transitivity structure.
Absolute clauses as hypotactic clauses can be considered as circumstances of the
primary clauses. Since the subject–predicate constructions realizing circumstances
can be seen as absolute clauses, then can subject–predicate constructions realizing
participants also be considered as absolute clauses? In clause complexes, depend-
ent clauses of expansion can form absolute clauses, then can dependent clauses
of projection form absolute clauses? Traditional grammar distinguishes three

© Springer-Verlag Berlin Heidelberg 2015 53


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_4
54 4  Research Design

relation types of absolute clauses: accompanying, subordination, and explanation.


Although the former two relations can be included in the category of adjuncts,
absolute clauses of explanation do not function as adjuncts at all, rather they func-
tion as appositives. According to the identification criteria of absolute clauses
developed in the last chapter, “nominal group + non-finite element” constructions
functioning as subject or complement can also form absolute clauses. On the other
hand, both subject and complement are participants of the main verb. If embed-
ded non-finite clauses can form absolute clauses, then can absolute clauses realize
parataxis? The three-level functional syntactic analysis provides us not only with
a new perspective to observe absolute clauses, but also with a basis for identify-
ing syntactic functions of absolute clauses. Therefore, the first question that this
research aims to answer is:
1. What types of relationship can absolute clauses realize?
It is generally acknowledged that language has been changing in the process of
historical development. The research of absolute clauses in traditional g­ rammar
includes such aspects as function types, case choice, historical distribution,
and stylistic distribution, for example, the origination of absolute clauses, the
application of absolute clauses in Old English, Middle English, and Modern
English. Modern grammarians, such as Jespersen (1933: 313, 1949: 62) and
Quirk et al. (1985: 1063), tend to believe that absolute clauses are formal and
unusual. Then, with the development of language, are absolute clauses gradu-
ally decreasing or increasing in use? Research shows that non-finite clauses
are “becoming more important in the recent history of English” (Leech et al.
2009: 2). Then, are the non-finite clauses spreading at the expense of finite
alternatives? Are absolute clauses as non-finite clauses spreading out of the
same r­eason? To answer these questions, it is necessary to resort to the real
language corpus to conduct synchronic and diachronic research of absolute
clauses and to explore their distributions. Therefore, the second question this
research aims to answer is:
2. What are the synchronic and diachronic distributions of absolute clauses?

4.2 Research Methods

This research adopts a combined method of qualitative research and corpus-based


quantitative research. This is also a constant method in SFL research. SFL ­studies
language-in-use. The theoretical systems are formulated with the data collected
from spoken and written discourses. Theories are then tested and improved in prac-
tical use. Halliday followed the field survey and real data analysis by Malinowski
and Wang, with the reference to the studies of Malinowski (1935) and Bühler
(1934) and worked out the theory of metafunction and the three-dimensional con-
text theory of genre, register and language (Martin 1992). These are all the theo-
retical abstraction on the basis of corpus analyses and qualitative descriptions.
4.2  Research Methods 55

For example, the research of cohesion patterns by Halliday and Hasan (1976) is a
model of combination of qualitative and quantitative research.
Corpus-based qualitative and quantitative linguistics research is becoming a
mainstream paradigm in the study of languages. Qualitative research is the basis
of quantitative research, and quantitative research makes qualitative research more
accurate. Quantitative and qualitative analyses are complementary to each other
(Bunge 1995). In actual studies, qualitative and quantitative methods are always
used with each other in order to accurately qualify on the basis of quantifying.
SFL puts emphasis on the complementarity between qualitative research and
quantitative research in both theoretical investigation and specific practice. The
discussion of function types and the development of criteria of identifying abso-
lute clauses in the framework of SFL should be carried out on the basis of qual-
itative description and quantitative analysis. The qualitative research of absolute
clauses provides a theoretical basis for the quantitative research.
“Quantitative research is a means for testing objective theories by examining
the relationship among variables” (Creswell 1994: 1). The quantitative research in
the framework of SFL is a realization of the notion of probability. Corpus is an
essential tool for carrying out quantitative research. Corpora help to count up the
overall occurrences of absolute clauses, the proportion of each type of function
realized by absolute clauses, and the occurrences of each function type of abso-
lute clauses. It is also helpful for counting up which types of function of absolute
clauses can or cannot be introduced by with and their probabilities of occurring.
Historical corpora can help to analyze the diachronic changes of the probabilities
and to explain the historical evolution of absolute clauses. The change of the abso-
lute clauses introduced by with can be used to test whether the independence of
absolute clauses in clause complexes is increasing or decreasing, and to further
test whether the use of clause complexes is increasing or decreasing. Synchronic
corpora can be helpful to analyze the distributions of absolute clauses in different
stylistic works and further to summarize and explain the stylistic distributions of
absolute clauses.
“SFL research is corpus-based” (Butler 1985: 15). “Two key tools of empirical
linguistics at the turn of the century are the corpus and the computer” (Sampson
2001: 12). “There are, therefore, close ties between corpus linguistics and SFL”
(Neale 2006) and “corpus-based methodology and text-based research have played
a central role in SFL since the beginning” (Matthiessen 2006). Corpus-based
research ties the form and meaning of language closer in certain context, and cor-
pus itself organically links the form, meaning and function of language together.
Halliday and Matthiessen (2004: 34–35) enumerate three plusses relating to the
use of the corpus: First, its data are authentic; second, its data include spoken lan-
guage; third, the corpus makes it possible to study grammar in quantitative terms.
Currently, “SFL is an ‘extravagant’ theory, which consciously provides a rich
description. This helps to explain why SFL corpus-based work is generally slow,
unmechanised and small-scale in comparison with corpus linguistics” (Thompson
and Hunston 2006). The basic situation is that there is “so much theory built over-
head with so little data to support it” (Halliday 1996[2002]). Although there are
56 4  Research Design

some urgent problems to be dealt with, corpus-based SFL research is becoming


increasingly significant.
Specifically, this research is a combination of top-down and bottom-up operations.
1. Top-down
The basic concepts of SFL form an abstract framework or model, with which
we can propose hypotheses, select systems, and determine research methods,
etc. In the relevant SFL studies, this kind of top-down quantitative research
is quite common. Absolute clauses are hypotactic non-finite clauses in clause
complexes, and the study of absolute clauses should start from the basic theo-
retical framework of clause complexes to propose hypotheses, test hypotheses
using the data retrieved from corpus, and summarize the general characteristics
of absolute clauses.
2. Bottom-up
The bottom-up research against the top-down research is a method focusing
on the reasoning, modifying, and accomplishing of hypotheses, theories, and
conclusions starting from language-in-use. Although absolute clauses are non-
finite clauses, different syntactic structures have inevitably different features
and functions. This research will test the basic functional syntactic theory of
SFL through real language analyses, rethink the basic theory in terms of the
basic features of absolute clauses and thereby pursue to revise and enrich this
theory framework.
In addition, this research also uses the method of comparison and analogy. The
selected diachronic and synchronic data tagged using TreeTagger 2.0, retrieved
using AntConc 3.2.4w, and counted and analyzed using UAM CorpusTool 2.8.12.
The classification of data is a combination of manual and automatic operations,
to ensure the authenticity and reliability of the data. The reason why the AntConc
3.2.4w software is used is that it supports regular expressions and makes retrieving
easy. In the preparation of regular expressions, we refer to the regular expression
compiling tool PatternBuilder.

4.3 Data Collection

At the second half of the twentieth century, linguistic research began to use intu-
itively invented sentences as the theoretical evidence. This is a reflection of the
subjectivity of the researchers, and hence cannot reveal the actual status of lan-
guage. SFL takes language as a social semiotic and holds that the meaning of
language comes from its use. Following this notion, except for a small number
of directly cited example sentences, all other example sentences used in this
research are actual data retrieved from such corpora as the Brown Family Corpora,
the BNC, COCA, and COHA.
The original purpose of the Brown Family Corpora is to facilitate the comparative
study of British English and American English. The six corpora in the Brown Family
4.3  Data Collection 57

American English British English

1961 Brown 1961 LOB 1961

1991/2 Frown 1992 FLOB 1991

2009 Crown 2009 CLOB 2009

Fig. 4.1  Six matching corpora in Brown Family

Fig. 4.2  Styles and number of texts in each corpus of Brown Family

Corpora are the Brown Corpus (1961), the LOB Corpus (1961), the Frown Corpus
(1992), the FLOB Corpus (1991), the Crown Corpus (2009), and the CLOB Corpus
(2009) (Fig. 4.1).
These six corpora cover a span of time of nearly 50 years. The strength of these
six corpora lies in their comparability: The fact that they are constructed according
to the same design, having virtually the same size and the same selection of texts
and registers represented by 500 matching text samples of 2,000 words each, total-
ing 1 million words. This means that we can use the Brown Family Corpora as an
effective tool for tracking the differences between written English in 1961, 1991,
1992, and 2009 (Fig. 4.2; Table 4.1).
We use the BNC and COCA because both corpora have a large vocabulary
(100 million and 460 million words, respectively). This allows enough data for
constructions of relatively lower occurring frequencies. The two corpora cover
a wide range of registers, including not only newspaper, popular magazines,
­academic, and fiction, but also spoken. They can facilitate the comparative study
of British English and American English and the stylistic distribution study. They
58 4  Research Design

Table 4.1  Styles and number of texts in the Brown Family Corpora


Group Category Content of category No. of
texts
Press A Reportage 44
B Editorial 27
C Review 17
Subtotal 88
General prose D Religion 17
E Skills, trades and hobbies 36
F Popular lore 48
G Belles lettres, biographies, essays 75
H Miscellaneous 30
Subtotal 206
Learned J Science 80
Subtotal 80
Fiction K General fiction 29
L Mystery and detective fiction 24
M Science fiction 6
N Adventure and Western 29
P Romance and love story 29
R Humor 9
Subtotal 126
Overall total for each corpus 500

also provide convenient search tools. The overall occurring frequency of absolute
clauses is relatively low, and different function types of absolute clauses have dif-
ferent occurring frequencies. As for some types of absolute clauses of lower fre-
quencies, such as those with personal pronoun subject, it is not likely to retrieve
enough occurrences to reflect the distribution trend from the Brown Family
Corpora of only 6 million words. However, the occurrences retrieved from so big a
corpus as COCA are enough to reflect the basic distribution of the retrieved item.
COCA has been expanded by 20 million words each year, since the early
1990s. To the June of 2012, COCA has a total number of more than 464 mil-
lion words. Table 4.2 shows that the BNC is 10 % spoken versus 90 % written,
while in COCA, the corpus is nearly evenly divided among such registers as spo-
ken, fiction, popular magazines, newspaper, and academic (20 % in each). For the
research of absolute clauses, the stylistic distribution of absolute clauses can be
counted from the Brown Family Corpora, and their distribution in spoken versus
written registers can be counted from COCA and the BNC (Table 4.3).
Absolute clauses are an ancient syntactic structure. It has been changing with
the evolution of language from Old English through Modern English. Since the
4.3  Data Collection 59

Table 4.2  Number of words of the Brown Family Corpora


Corpus Register
Press General prose Learned Fiction Total
Brown 181,085 423,160 163,309 259,467 1,027,021
Frown 181,748 421,933 163,228 260,414 1,027,323
Crown 180,980 422,799 163,197 259,250 1,026,226
Lob 179,604 418,137 162,322 258,722 1,018,785
Flob 180,703 419,990 163,286 260,664 1,024,643
Clob 179,680 421,163 163,139 259,484 1,023,466
Total 1,083,800 2,527,182 978,481 1,558,001 6,147,464

Table 4.3  COCA and BNC compared in terms of register balance and number of words
REGISTER COCA (millions of words) BNC (millions of words)
Spoken 95 10
Fiction 90 17
Popular magazines 96 16
Newspaper 92 11
Academic 91 16
Other 30
Total 464 100

nineteenth century, grammarians have paid enough attention to this structure and
its evolution. Diachronic corpora are an effective tool for the research of the evo-
lution of language. The Brown Family Corpora cover a relatively short span of
time. Although the current application and distribution of absolute clauses can be
acquired, it is not possible to catch the diachronic evolution of absolute clauses in
such a short history as less than 50 years. To address this deficiency, we choose
to use COHA (the Corpus of Historical American English) to conduct diachronic
analyses of absolute clauses. COHA is the largest structured corpus of historical
English with a total number of words of 400 million, covering a span of time from
1810 to 2009. The corpus is balanced by register across the decades. For e­ xample,
fiction accounts for 48–55 % of the total in each decade (1810s–2000s), and the
corpus is balanced across decades for sub-registers and domains as well (e.g., the
sub-registers of prose, poetry, and drama in the register of fiction). This balance
across registers and sub-registers allows researchers to examine changes of lan-
guage and be reasonably certain that the data reflect actual changes in the “real
world,” rather than just being artifacts of a changing register balance. As for the
research of absolute clauses, this balance helps to examine their changes of fre-
quencies and stylistic distributions during the 200 years of language evolution
(Table 4.4).
60 4  Research Design

Table 4.4  Registers, number of words, and historical distribution in COHA


Decade Fiction Popular magazines Newspapers Non-fiction books Total
1810s 641,164 88,316 0 451,542 1,181,022
1820s 3,751,204 1,714,789 0 1,461,012 6,927,005
1830s 7,590,350 3,145,575 0 3,038,062 13,773,987
1840s 8,850,886 3,554,534 0 3,641,434 16,046,854
1850s 9,094,346 4,220,558 0 3,178,922 16,493,826
1860s 9,450,562 4,437,941 262,198 2,974,401 17,125,102
1870s 10,291,968 4,452,192 1,030,560 2,835,440 18,610,160
1880s 11,215,065 4,481,568 1,355,456 3,820,766 20,872,855
1890s 11,212,219 4,679,486 1,383,948 3,907,730 21,183,383
1900s 12,029,439 5,062,650 1,433,576 4,015,567 22,541,232
1910s 11,935,701 5,694,710 1,489,942 3,534,899 22,655,252
1920s 12,539,681 5,841,678 3,552,699 3,698,353 25,632,411
1930s 11,876,996 5,910,095 3,545,527 3,080,629 24,413,247
1940s 11,946,743 5,644,216 3,497,509 3,056,010 24,144,478
1950s 11,986,437 5,796,823 3,522,545 3,092,375 24,398,180
1960s 11,578,880 5,803,276 3,404,244 3,141,582 23,927,982
1970s 11,626,911 5,755,537 3,383,924 3,002,933 23,769,305
1980s 12,152,603 5,804,320 4,113,254 3,108,775 25,178,952
1990s 13,272,162 7,440,305 4,060,570 3,104,303 27,877,340
2000s 14,590,078 7,678,830 4,088,704 3,121,839 29,479,451
Total 207,633,395 97,207,399 40,124,656 61,266,574 406,232,024

4.4 Data Processing

We first classify the six Brown Family Corpora according to the four registers of
press, general prose, learned, and fiction into 24 sub-corpora, tag all the sub-cor-
pora using TreeTagger 2.0 and write the following regular expression according to
the formal types of absolute clauses.
RE1.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|N\w+)\s(\
S+_RB\w*\s)*(\S+_TO\s\S+_V[BDHV]|\S+_(V(B|D|H|V)[GN]|JJ\w*|RB\
w*|IN|DT|N\w+))(\S+_IN\s)*(\S+_(DT|CD|P\w*)\s)*(\S+_(PP|N\w+)\s)*
Then, we use AntConc 3.2.4w for corpus search. The retrieved occurrences
match all the constructions consisting of a nominal group or personal pronoun and
a non-finite element or a verbless element. Finally, we check all the concordance
lines, exclude all the lines that cannot form absolute clauses, and extract all the
absolute clauses. However, we retrieved a total of 95,730 concordance lines using
the above regular expression. It might be a hard job to extract absolute clauses
manually from these nearly 100 thousand concordance lines, and in these lines,
only a very few constructions can form absolute clauses; thereby, it is necessary
4.4  Data Processing 61

to limit the searching conditions to some extent to ensure a relatively higher rate
of absolute clauses in the concordance lines. For example, we can search the con-
structions consisting of a nominal group and a non-finite element directly ending
the structure. Therefore, we write the following regular expressions, respectively,
according to the formal types of absolute clauses.
RE2.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|N\w+)\s(\
S+_RB\w*\s)*\S+_V(B|D|H|V)[GN]\s(\S+_IN\s)*(\S+_(DT|CD|P\w*)\s)*(\
S+_(PP|N\w+)\s)*\S+_(,|SENT|:)
RE3.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|N\w+)\s\
S+_(JJ\w*|RB\w*)\s\S+_(,|SENT|:)
RE4.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|N\w+)\s\
S+_TO\s\S+_V[BDHV]\s(\S+_(DT|CD|P\w*)\s)*(\S+_(PP|N\w+)\s)*\
S+_(,|SENT|:)
RE5.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|N\w+)\s\
S+_IN\s(\S+_(DT|CD|P\w*)\s)*\S+_N\w+\s\S+_(,|SENT|:)
RE6.\S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)*\S+_(PP|NP\w*)\s(\
S+_(DT|CD|P\w*)\s)\S+_N\w+\s\S+_(,|SENT|:)
The proportion of absolute clauses retrieved from the concordance lines searched
using the restricted regular expressions has substantially increased. We retrieved 817
collocations consisting of a nominal group or pronoun and a present or past partici-
ple from the second regular expression, 1,165 consisting of a nominal group or pro-
noun and an adjective or adverb from the third, 105 consisting of a nominal group
or pronoun and an infinitive from the fourth, 2,703 consisting of a nominal group
or pronoun and a prepositional phrase from the fifth and 417 consisting of a proper
noun or personal pronoun and a nominal group from the sixth. Finally, after exclud-
ing the collocations not being able to form absolute clauses, we got 221 absolute
clauses from the second regular expression, 89 from the third, 3 from the fourth, 66
from the fifth, and 3 from the sixth.
In addition, it being … and there being … are two common types of absolute
clauses. By direct word search, we retrieved 21 collocations of it being … and 29
of there being …, among which there are 6 and 13 absolute clauses, respectively.
Based on the absolute clauses retrieved from the above regular expressions, this
research uses UAM CorpusTool 2.8.12 to carry out comparative analyses on the
distributions of absolute clauses in the Brown Family Corpora, including the dis-
tributions of overall occurrences, region, register and case, etc.
It is convenient to carry out comparative study using the Brown Family
Corpora. However, because of the small number of words and short span of time,
we extracted only a small number of absolute clauses from the Brown Family
Corpora. These occurrences may not be enough to reflect the general character-
istics of absolute clauses. Moreover, these corpora lack the spoken register rep-
resenting the informal style. Therefore, we choose to use the BNC and COHA.
The BNC includes such registers as spoken, newspaper, popular magazines, fic-
tion, academic, and others. The registers of newspaper, popular magazines, fiction,
and academic are roughly equivalent to the classification of registers. However, the
62 4  Research Design

register of popular magazine includes only pop lore in the BNC, while the register
of general prose in the Brown Family Corpora consists of religion, trades, hobbies,
popular lore, belles lettres, biographies, essays and miscellaneous, etc. These con-
tents are included in the registers of miscellaneous and non-academic, and thereby
we choose popular magazine, non-academic, and miscellaneous in the BNC as the
register of general prose. More importantly, the BNC corpus includes the spoken
register representing the informal style.
The BNC is available online, so there is no way to write regular expressions,
or to search using AntConc 3.2.4w. However, we can write the following search
query in terms of the CLAWS7 Tagset.
SQ1. [y*] [nn*]|[p*] [v?n*]|[v?g*]|[nn*]|[j*]|[r*]|[i*]
With this search query, we can retrieve all the collocations consisting of a noun
or pronoun and a present or past participle, noun, adjective, adverb or preposi-
tional phrase following any punctuation. However, the number of concordance
lines is so big that it is quite difficult to extract all the absolute clauses by hand, so
we make appropriate limits on the search query. For example, we can search out
the collocations directly ended with an adjective after a noun or pronoun, or the
collocations directly ended with a prepositional phrase composed of a preposition
and a noun. Therefore, we write the following search queries according to the for-
mal types of absolute clauses.
SQ2. [y*] [nn*]|[p*] [v?g*] (0-2) [y*]
SQ3. [y*] [nn*]|[p*] [v?n*] (0-2) [y*]
SQ4. [y*] [nn*]|[p*] [j*] [y*]
SQ5. [y*] [nn*]|[p*] [i*] [nn*] [y*]
Using the above four search queries, we can retrieve the absolute clauses com-
posed of a noun or pronoun and a present or past participle, adjective and prep-
ositional phrase. The reason why we choose these four types of collocation is
that within all forms of absolute clauses, these four types are of the highest fre-
quency of occurrence and are of a higher probability to form absolute clauses (See
Table 6.1 and Fig. 6.1).
Based on occurrences of absolute clauses retrieved from the above search que-
ries, this research analyzes the stylistic distribution of absolute clauses in the BNC
and carries out a comparative analysis with the research result based on the Brown
Family Corpora.
COHA covers a span of time from 1810 to 2010. We count the absolute clauses
across the decades. Like the BNC, we write the following search query according
to the CLAWS7 Tagset.
SQ1. [y*] [nn*]|[p*] [v?n*]|[v?g*]|[nn*]|[j*]|[r*]|[i*]
COHA has a total number of words of 400 million. It is a hard work to extract
all the absolute clauses from the collocations retrieved from so big a corpus by
hand, so we make some limits on the search query. We only search absolute
4.4  Data Processing 63

clauses composed of a noun or pronoun and a present participle. Therefore, we


write the following search queries as given below.
SQ2. [y*] [nn*]|[p*] [v?g*] (0-2) [y*]
SQ3. [y*] [at*] [nn*]|[p*] [v?g*] (0-2) [y*]
The second search query can be described as a construction consisting of a
noun or pronoun and a present participle with zero to two words followed after a
punctuation mark. The third search query can be described as a construction con-
sisting of an article preceding a noun or pronoun and a present participle with zero
to two words followed after a punctuation mark. Absolute clauses can be extracted
from all the retrieved concordance lines fulfilling the search queries by hand.
Based on occurrences of absolute clauses retrieved from the above search
queries, this research analyzes the diachronic distribution of absolute clauses in
COHA, and carries out a comparative analysis with the research result based on
the Brown Family Corpora.
This research is a corpus-based quantitative and qualitative research. The
­qualitative analysis of the formal and functional types of absolute clauses will be
combined with the quantitative analysis of the actually used data in the corpus.
The quantitative analysis will dominate in the discussion of the historical evolution
and stylistic distribution.

References

Bühler, K. (1934). Sprachtheorie: Die darstellungfunktion der sprache. Jena: G. Fischer.


Bunge, M. (1995). Quality, quantity, pseudo quantity and measurement in social science. Journal
of Quantitative Linguistics, 2(1), 1–10.
Butler, C. S. (1985). Systemic linguistics: Theory and application. London: Batsford.
Creswell, J. W. (1994). Research design: Qualitative and quantitative approaches. Thousand
Oaks, CA: Sage Publications.
Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. London: Longman Group.
Halliday, M. A. K. (1996[2002]). On grammar and grammatics. In J. Webster (Ed.), Collected
works of M. A. K. Halliday, vol. 1: On grammar (pp. 384–417). London: Continuum.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2004). An introduction to functional grammar
(3rd ed.). London: Edward Arnold.
Jespersen, O. (1933). Essentials of English grammar. London: Allen & Unwin.
Jespersen, O. (1949). A modern English grammar on historical principles. London: Allen &
Unwin.
Leech, G., Hundt, M., Mari, C., & Smith, N. (2009). Change in contemporary English: A gram-
matical study. Cambridge: Cambridge University Press.
Malinowski, B. (1935). The language of magic and garden, vol. 2: Coral gardens and their
magic. New York: American Book Co.
Martin, J. R. (1992). English text: System and structure. Philadelphia, Amsterdam: John
Benjamins.
Matthiessen, C. M. I. M. (2006). Frequency profiles of some basic grammatical systems: An
interim report. In G. Thompson & S. Hunston (Eds.), System and corpus: Exploring connec-
tions (pp. 103–142). London: Equinox.
64 4  Research Design

Neale, A. (2006). Matching corpus data and system networks: Using corpora to modify and
extend the system networks for transitivity in English. In G. Thompson & S. Hunston (Eds.),
System and corpus: Exploring connections (pp. 143–163). London: Equinox.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London, New York: Longman.
Sampson, G. (2001). Empirical linguistics. London, New York: Continuum.
Thompson, G., & Hunston, S. (2006). System and corpus: Two traditions with a common
ground. In G. Thompson & S. Hunston (Eds.), System and corpus: Exploring connections
(pp. 1–14). London: Equinox.
Chapter 5
Relationships Realized by Absolute Clauses

5.1 Introduction

Traditional grammar distinguishes three types of subordinate clauses: adverbial


clauses, nominal clauses (subject clauses, object clauses, and appositive clauses),
and adjectival clauses (attributive clauses). Although the three types of subordi-
nate clauses can all be non-finite, only adverbial clauses have the potential to form
absolute clauses. However, the traditional definition of absolute clauses as clauses
is not clear, because absolute clauses can function as clausal adjuncts of time,
cause, condition, concession, etc., but there are no adverbial clauses of attendant
circumstance. Furthermore, the explanation of with in augmented absolutes in tra-
ditional grammar is also vague. It only mentions that with introduces the subject of
the absolute construction, but never explains in detail why the subject is introduced
by with nor the syntactic or semantic role of with. “A general definition would
involve both grammatical and semantic considerations.” (Halliday and Matthiessen
2004: 51). Therefore, the traditional definition of absolute clauses includes both
their structure types and function types. SFL does not confine the functions real-
ized by a lexico-grammatical item. One function can be realized by more than one
forms and one form can realize more than one functions. This opens up a large
relation potential for absolute clauses. “The class of an item indicates in a general
way its potential range of grammatical function.” (ibid.: 52) The SFL classification
of relation types of clause complexes is different from that of traditional grammar.
SFL construes a complicated relationship system network between the two clauses
constituting a clause complex along two dimensions: interdependency and logico-
semantic relation. Since absolute clauses are an alternative form of finite clauses,
they have a similar meaning potential to finite clauses.

© Springer-Verlag Berlin Heidelberg 2015 65


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_5
66 5  Relationships Realized by Absolute Clauses

5.2 Relation Potential of Absolute Clauses

SFL attributes absolute clauses to the category of non-finite clauses. Thus, abso-
lute clauses have the same relation potential as non-finite clauses.

5.2.1 Relation System Network

In Chap. 2, we discussed the formal and functional types of absolute clauses. The
types of absolute clauses in traditional grammar are represented in Fig. 5.1.
The syntactic analysis of SFL is meaning-based. The meaning-based syntac-
tic analysis facilitates the discussion of covert categories of grammar. Its focus of
research is not on the overt markers. For example, SFL does not distinguish the cases
of nouns or pronouns, taking case as a covert category, but distinguishes such rank
units as clause, group/phrase, word, and morpheme at the lexico-grammatical level.
Absolute clauses are composed of a nominal group and a non-finite verbal group,
nominal group, adverbial group or prepositional phrase at the lexico-grammatical
level. They have the same transitivity structure, thematic structure, and informational
structure as finite clauses at the clause rank. What differs absolute clauses from finite
clauses in structure is the lack of Finite in the mood structure. The SFL-based formal
types of absolute clauses are shown in Fig. 5.2.

Fig. 5.1  Type system network of absolute clauses in traditional grammar


5.2  Relation Potential of Absolute Clauses 67

noun
NOMINAL
GROUP
pronoun
present
absolute clause participle
nonfinite past
infinitive
VERBAL noun
GROUP
nominal
adverbial adjective
verbless
prepositional

Fig. 5.2  Type system network of absolute clauses in SFL

As can be seen, absolute clauses are composed of nominal groups and non-finite
verbal groups in both traditional grammar and SFL. The formal difference between
the two lies in that SFL has no case requirement for the nominal group function-
ing as subject. In addition, SFL does not define the functions realized by absolute
clauses. This opens up a wide potential of functions for absolute clauses.

5.2.2 Relation Potential

According to SFL, absolute clauses are non-finite clauses with an explicit subject
but with no finite or explicit conjunctive expressions. The functions of absolute
clauses are not limited to the traditional clausal adjuncts and attendant circum-
stances. Since absolute clauses are at the clause rank, they have the potential to
realize multiple meanings. First, absolute clauses can enhance or extend the pri-
mary clauses as hypotactic clauses. Hypotactic enhancing clauses are adverbial
clauses in traditional grammar, while attendant circumstances in traditional gram-
mar are hypotactic extension in SFL. When defining lexico-grammatical items,
SFL does not define their functions. Therefore, besides hypotactic expansion,
absolute clauses have the potential to realize other semantic relations. For exam-
ple, absolute clauses can be used independently, and the logico-semantic relations
are realized through textual cohesive devices but not structural devices. Absolute
clauses have also the potential to realize parataxis and hypotactic projection in
clause complexes and to realize participants, or the modifier or appositive of par-
ticipants in simple clauses. This relation potential is represented in Fig. 5.3.
Absolute clauses have the potential to realize all the relationships shown in
Fig.  5.3. In the next section, we will analyze in detail all the relationships that
absolute clauses can actually realize.
68 5  Relationships Realized by Absolute Clauses

parataxis
TAXIS
hypotaxis
idea
secondary projection
locution
LOGICO-
SEMANTIC elaboration
expansion extension
absolute clause independent enhancement

subject
one level
complement
embedded
appositive
two level
modifier

Fig. 5.3  Relation potential system of absolute clauses

5.3 Dependent Absolute Clauses

The secondary clauses realizing hypotaxis are referred to as dependent clauses.


Logico-semantic relations include expansion and projection. Absolute clauses as
non-finite clauses function as dependent clauses, realizing hypotactic expansion
and projection.

5.3.1 Expansion

5.3.1.1 Elaboration

In a clause complex of elaboration, “one clause elaborates on the meaning of


another by further specifying or describing it… The secondary clause does not
introduce a new element into the picture but rather provides a further characteri-
zation of one that is already there, restating it, clarifying it, refining it or adding
a descriptive attribute or comment.” (Halliday and Matthiessen 2004: 396). The
hypotactic clause realizing elaboration is a nonrestrictive relative clause which
is a descriptive explanation of the primary clause or part of the primary clause.
Elaboration can be paratactic or hypotactic. Parataxis and hypotaxis are different
in meaning and realizing manners, but they have a tone concord. That is to say, the
two clauses in an elaborative clause complex both have a tone group, and the two
5.3  Dependent Absolute Clauses 69

tones are the same. The difference between hypotactic and paratactic elaborations
is that the former are relative clauses introduced by relative pronouns or relative
adverbs functioning as both structural Theme and topic Theme, which correspond
to the nonrestrictive attributive clauses in traditional grammar, while the latter usu-
ally have no conjunctive expressions realizing elaboration.
Paratactic and hypotactic elaborations are complementary to each other in
meaning. The former involves exposition, exemplification, and clarification and
the latter, description. For example,
5-1a. She wasn’t a show dog; I didn’t buy her as a show dog. (Halliday and
Matthiessen 2004: 398) (exposition)
b. You’re too old for that game; you couldn’t bend over (ibid.: 398)
(exemplification)
c. They weren’t show animals; we just had them as pets. (ibid.: 399)
(clarification)
d. You followed them with The Greenlanders, which seems to me more ambi-
tious. (ibid.: 399) (description)
The two clauses in a paratactic clause complex are both finite. They do not need
conjunctions to realize the relator. If there are conjunctions, they are ­cohesive rather
than structural markers. For example, in exposition, the secondary clause repeats
more or less the information of the primary clause through lexical ­repetition or
synonymy to function as emphasis. In exemplification, the ­secondary clause main-
tains the lexical cohesive relation with the primary clause through hyponymy or
meronymy. In clarification, the secondary clause clarifies the thesis of the primary
clause through synonymy. Hypotactic elaborative clauses are nonrestrictive relative
clauses, a descriptive explanation of a part of or the whole primary clause. The log-
ico-semantic relations between the primary and the secondary clauses are realized
by relative pronouns or relative adverbs. Non-defining relative clauses and defining
relative clauses embedded in nominal groups functioning as qualifier are different
both in meaning and in form. The difference in meaning is that defining relative
clauses define a specific subset, while non-defining relative clauses add new fea-
tures to the already-specific clause or a constituent of a clause. The difference in
form is that the non-defining relative clauses are separated from the primary clauses
with a punctuation mark, while the defining relative clauses usually follow the ante-
cedent immediately.
The semantic and formal differences can be used to answer the following two
questions: (1) Why can the non-subject relatives of defining relative clauses be
omitted? (2) Why cannot the relatives of non-defining relative clauses be that?
The answer to the first question is as follows. When the relative is not the
subject of a relative clause, the subject of the relative clause is not co-referential
with the subject of the clause within which it is embedded. The logico-semantic
relation of the two clauses is clear, so there may be no overt relatives. When the
relative is the subject, the omission of the relative will result in the antecedent
playing a syntactic role in both the relative clause and the main clause. This is
ungrammatical.
70 5  Relationships Realized by Absolute Clauses

The following is the answer to the second question. Since non-defining relative
clauses functions as the explanation but not the restriction of the antecedent, the
relationship between them is relatively loose. If that is used as the subject of the
relative clause, it may be misinterpreted as a demonstrative pronoun functioning as
the subject of a paratactic clause.
The explanations of the two questions show that the rank status of non-defining
relative clauses, that is, the elaborative dependent clauses, is ambiguous. It should
be in between the embedded clauses functioning as modifier and paratactic elabo-
rative clauses, the three forming a cline.
The elaborative secondary clause can also be non-finite. The meaning of non-
finite clauses is usually not as specific as that of their finite equivalents. Their
semantic domain can be a nominal group or a larger syntagma, or even a whole
clause. Elaborative non-finite clauses do not need prepositions or conjunctions to
realize relators, because “relators themselves embody meanings of extension or
enhancement” (Yang 2003: 79). Non-finite dependent clauses may have their finite
equivalents, as in 5-2, or not, as in 5-3.
5-2a. He was an absolute loner of a man, pursuing some dream of exploration in
the jungles. (Halliday and Matthiessen 2004: 403)
b. He was an absolute loner of a man, who pursued some dream of exploration
in the jungles.
5-3 I worked for a local firm at that time, selling office equipment. (ibid.: 404)
Like continuing clauses, the elaborative relation realized by these non-finite
dependent clauses is clarification. In this case, non-finite clauses may have an
explicit subject, as in 5-4a. If the subject is not introduced by a preposition, they
may form absolute clauses, as in 5-4b.
5-4a. It’s a much bigger house, for the children to have their own rooms.
(Halliday and Matthiessen 2004: 403)
b. John went off by himself; the rest of us staying behind. (ibid.: 404)
As is mentioned above, attributive clauses in traditional grammar include nonre-
strictive and restrictive attributive clauses. The antecedent of a restrictive attributive
clause is a nominal group. Attributive clauses function as the modifier of ante-
cedents. Restrictive attributive clauses are embedded clauses in SFL, rankshifted
to function as modifiers in the clauses in which they are embedded. They do not
change the syntactic structure of the clauses in which they are embedded, and the
whole structure is still a simple clause. The antecedent of a nonrestrictive clause can
be a nominal group or the whole primary clause. Nonrestrictive clauses augment the
original clause structure, and the whole construction is a hypotactic clause complex.
5-5a. She opened a bakery, which proved to be an instant success. (COCA_FIC)
b. He had a theory, which was not original. (COCA_FIC)
c. He had a theory, not original.
The non-defining relative clause which proved to be an instant success in
5-5a functions to explain the primary clause She opened a bakery as antecedent. The
5.3  Dependent Absolute Clauses 71

relative clause in 5-5b which was not original is a relational clause, in which not orig-
inal is the attribute of the antecedent a theory. In this case, the finite clause can be
changed into the non-finite form, with the subject and finite carrying no information
omitted, leaving only the attribute carrying information, as in 5-5c. If the attribute
itself is composed of a nominal group and a non-finite verbal group forming a non-
finite clause with subject, this construction forms an absolute clause. For example,
5-6a. I just wondered why I heard a noise, the gerbils running. (BNC_SPOK)
b. The noise doesn’t stop—glass breaking, wood popping, everything smash-
ing down. (COHA_FIC)
c. Then, he was aware of another presence, a kigrin circling him, wading
through the murky water. (COHA_FIC)
Each of the absolute clauses in 5-6 explains a nominal group in the primary
clause, realizing elaboration.

5.3.1.2 Extension

In a clause complex, extension means that “one clause extends the meaning of
another by adding something new to it” (Halliday and Matthiessen 2004: 405).
Three subtypes of paratactic extension can be recognized: addition, variation, and
alternation. Addition can further be divided into positive addition (and), negative
addition (nor) and adversative (but), and variation can be divided into replacive
(instead) and subtractive (except). Hypotactic extension also embraces addition,
variation, and alternation, but the secondary clauses realizing extension are
not continuing clauses but dependent clauses. There is not a sharp line between
paratactic and hypotactic extensions. “If the extending clause could precede
(thereby becoming thematic in the clause complex), the relationship is hypotac-
tic” (Halliday and Matthiessen 2004: 408). These extensive relations are shown in
Fig. 5.4.
Non-finite extending clauses cover the subtypes of addition and variation, but
not alternation, and there is no negative addition in the subtype of addition. Thus,

Fig. 5.4  Extension system parataxis


network TAXIS
hypotaxis

positive
extension
addition negative
adversative
LOGICO-
alternation
SEMANTIC
replacive
variation
subtractive
72 5  Relationships Realized by Absolute Clauses

Fig. 5.5  Extension system ----


network of non-finite clauses TAXIS
hypotaxis

positive
extension
addition ----

adversative
LOGICO-
----
SEMANTIC
replacive
variation
subtractive

extension of non-finite clauses embraces positive addition, adversative addition,


replacive variation, and subtractive variation (see Fig. 5.5 and example 5-7).
5-7a. Jimi was my favourite singer, besides being my favourite guitarist.
(BNC_MAG) (addition: positive)
b. It gives a lovely natural glow to the wood, without making it appear too
shiny. (BNC_MISC) (addition: adversative)
5-8a. Instead of going into the green room, we went to the girls’ dressing room.
(BNC_MISC) (variation: replacive)
b. Our life has no meaning other than serving our controllers. (BNC_FIC)
(variation: subtractive)
The non-finite clauses in 5-7 and 5-8 realize extension of addition and exten-
sion of variation, respectively. Extension of variation always requires conjunctive
prepositions to realize relator. Non-finite clauses of variation with subject cannot
form absolute clauses. For example,
5-9a. Instead of payments ending next year, they will have to keep paying for
another year or two. (COCA_MAG) (variation: replacive)
b. But the fact of the matter is that there’s been no evidence other than people
alleging this and saying this. (COCA_SPOK) (variation: subtractive)
“With the additive and adversative, however, there may be no conjunctive
expression; such clauses are therefore identical with non-finite elaborating clauses,
except that in speech they are not marked by tone concord.” (ibid.: 410)
5-10a. She stooped, looked inside, then straightened, angry, talking to herself.
(COCA_FIC) (addition: positive)
b. Slowly, hardly knowing what she did, Alexandra nodded. (COCA_MAG)
(addition: adversative)
Non-finite additive and adversative clauses with subject can form absolute
clauses. Therefore, two types of absolute clauses of extension can be recognized:
positive addition and adversative addition (see Fig. 5.6).
5.3  Dependent Absolute Clauses 73

Fig. 5.6  Extension system ----


network of absolute clauses TAXIS
hypotaxis

positive
extension
addition ----
adversative
LOGICO- ----
SEMANTIC

----

At the same time, additive absolute clauses can be introduced by with or with-
out, forming augmented absolute clauses, implying “an accompanying circum-
stance to the situation described in the matrix clause” (Quirk et al. 1985: 1124).
For example,
5-11a. He leaned forward, (with) his hands on his knees. (COCA_FIC) (addition:
positive)
b. They had been working for 18 months, spending lots of money (with)no
income coming in. (COCA_MAG) (addition: adversative)

5.3.1.3 Enhancement

“In enhancement one clause (or sub-complex) enhances the meaning of another
by qualifying it in one of a number of possible ways: by reference to time, place,
manner, cause or condition.” (Halliday and Matthiessen 2004: 410) The two com-
monest ways of enhancing are time and cause. Paratactic enhancement is a kind
of coordination with a circumstantial feature realized by conjunctions or conjunc-
tional groups. Hypotactic enhancing clauses are known in traditional grammar as
adverbial clauses. Finite clauses of enhancement always need conjunctions to real-
ize the dependent relationship and the circumstantial relationship. Such conjunc-
tions are referred to as binders. For example,
5-12a. By the time I got outta that house, my mouth run dry and my knees wob-
bled. (COCA_FIC) (time)
b. House prices will tend to rise where the sun shines. (COHA_MAG) (place)
c. As it happens, railroad shopping is a growth industry in Zurich. (COCA_
NEWS) (manner)
d. We are reducing the numbers of people who qualify for free food because
people are cheating. (COCA_NEWS) (cause)
e. If I had a look, I’d be sure to find it. (BNC_MISC) (condition)
74 5  Relationships Realized by Absolute Clauses

Non-finite clauses of enhancement can also realize the relationships of time, man-
ner, cause, condition, and so on. Such conjunctive expressions as conjunctions,
prepositions, and zero conjunctions can all function as relators. For example,
5-13a. When returning home, the group was instructed to remain in silence and
(COCA_ACAD)
b. On returning home he sets to work. (COCA_MAG)
c. Returning home, Mick throws a party for the local poor. (COCA_FIC)
Non-finite clauses of time, manner, cause, or condition may have explicit con-
junctive expressions. For example,
5-14a. Create an extensive pros-and-cons list before making a decision.
(COCA_MAG) (time)
b. Ford hopes to beat Corker by making the race a referendum on the
Republican Congress. (COCA_NEWS) (manner)
c. I am no parenting expert, despite having four children. (COCA_MAG)
(concession)
d. But in real life, most people don’t eat fresh produce without doing some-
thing to it first. (COCA_MAG) (condition)
Non-finite clauses of time and cause (reason, purpose, and result) may have no
explicit conjunctive expressions. For example,
5-15a. Having finished her little speech, she turned gracefully and left the room.
(COCA_FIC) (time)
b. Being a Stone elemental, I could feel the vibrations in the rock
(COCA_FIC) (cause: reason)
c. To win the war, early detection of the deep and hidden cancers will be key.
(COCA_MAG) (cause: purpose)
d. The fish circled four times, then dove from sight, never to be seen again.
(COCA_FIC) (cause: result)
Like non-finite clauses of extension, the logical subject of non-finite clauses
of enhancement without explicit subject is co-referential with the subject of the
primary clauses. If the subject of non-finite clauses is not co-referential with
the subject of the primary clause, the subject of the dependent clause is always
the speaker, or the agent of the receptive clause, or even some certain unimpor-
tant nonspecific entity. However, such dependent clauses usually have their own
explicit subject. When this explicit subject is case marked, it may be oblique (e.g.,
him) or possessive (e.g., his). For example,
5-16a. In order for him to salvage his legacy, he’s going to have to really do what
he can to maintain some goodwill. (COCA_SPOK)
b. With him being down, everybody on the other team piled on him.
(COCA_FIC)
c. With his being on the road so much, he wanted Dusty to be able to defend
herself and her mother. (COCA_FIC)
5.3  Dependent Absolute Clauses 75

It is shown in 5-16 that the logical subject of perfective non-finite clauses is


oblique and that of imperfective non-finite clauses can be oblique or posses-
sive. If oblique, the non-finite element is a present participle; if possessive, ger-
und. Modern English tends to be oblique, indicating that such clauses in Modern
English are not rankshifted, but dependent clauses.
The circumstantial relations of non-finite dependent clauses are realized by con-
junctions or conjunctive prepositions. The meaning of relator of conjunctive prep-
ositions is relatively unspecific. The meaning of non-finite clauses introduced by
conjunctive prepositions depends on the meaning of the primary clauses. However,
this applies only to non-finite clauses of time, manner, and cause. For example,
5-17a. Without being aware of it, we were distracted by the smell of the skunk
and not by what the skunk had done. (COCA_FIC) (cause: reason)
b. Without being told, I knew the bluffs contained the Camp of the Dead.
(COCA_FIC) (condition: concession)
c. Without being ordered, the soldiers stepped up their pace. (COCA_FIC)
(unclear)
The four types of non-finite clauses of enhancement can all have their own sub-
ject, the relator being realized by conjunctive prepositions. For example,
5-18a. Before flags flying at half-mast, the teams traded touchdowns and were tied
at halftime. (COCA_MAG) (time)
b. The social studies are taught in this class through children taking turns
reading a textbook. (COCA_ACAD) (manner)
c. You may experience mood swings because of changes occurring.
(COCA_MAG) (cause)
d. He was worried that something might happen without him being there.
(COCA_FIC) (condition)
When the logico-semantic relations are not emphasized, these enhancing non-
finite clauses with subject may not necessarily have explicit conjunctive expres-
sions, hence forming absolute clauses. Because the meaning of non-finite clauses
introduced by conjunctive prepositions depends on the meaning of primary clauses
and the meaning of with as conjunctive preposition is nonspecific, with can be
added to all absolute clauses of enhancement, forming augmented absolute clauses
defined by traditional grammar. For example,
5-19a. (With) The dishes done, they returned to the living room. (COCA_FIC)
(time)
b. Then proceed down the line from oldest to youngest, (with) the children
taking turns putting Post-it notes on items they are claiming. (COCA_
NEWS) (manner)
c. (With) Funding being limited, it is not always possible to purchase foreign
maps with Scottish associations. (BNC_MISC) (cause)
d. (With) Weather permitting, they shuttle up and down in packs. (COCA_
NEWS) (condition)
76 5  Relationships Realized by Absolute Clauses

5.3.2 Projection

Absolute clauses as non-finite clauses can only realize hypotaxis. SFL distin-
guishes two types of hypotaxis: hypotactic expansion and hypotactic projection.
Now that absolute clauses can realize hypotactic expansion, then can they real-
ize hypotactic projection? According to the criteria of identifying absolute clauses
developed in Chap. 3, non-finite clauses of expansion can form expanding abso-
lute clauses, and non-finite clauses of projection can also form projecting absolute
clauses.
Traditional grammar labels absolute clauses with adjuncts. However, Curme
(1931: 157) identifies two classes of nominative absolutes according to func-
tion: absolute nominative in adverbial clauses and absolute nominative in subject
clauses. The subject of the absolute nominative in subject clauses is not structur-
ally related to the main verb of the main clause. For example,
5-20a. Three such rascals hanged in one day is good work for society. (Curme
1931: 157)
b. Women having the vote reduces man’s political power. (ibid)
c. Them trying to sing a song was just too horrible. (Reuland 1983: 101)
Seen from the transitivity structure, clauses functioning as both subject and
complement realize participants. Since clauses of subject can form absolute
clauses, clauses of complement can also form absolute clauses. Accordingly, this
kind of absolute clauses functioning as complement can be referred to as absolute
nominative in compliment clauses. For example,
5-21 We heard the gun firing several times. (BNC_FIC)
However, this view is in conflict with that of the traditional grammar, accord-
ing to which the gun in 5-21 functions as the object of heard and firing several
times as the complement of the object. According to SFL, the non-finite clause the
gun firing several times as a whole functions as the complement of heard. This
means that “functional syntactic analysis can better realize the meaning of clauses
than traditional syntactic analysis” (Huang 1998b), because what we heard is the
sound of the gun, but not the gun itself. In fact, the gun firing several times in 5-21
is a subject–predicate construction. The subject the gun is not directly related to
the verbal group heard, and there is no explicit conjunction (complementizer) that
between them. If the subject clause in 5-20 is an absolute clause, then the comple-
ment clause in 5-21 can also be seen as an absolute clause. It is certain that not all
“nominal group + participial phrase” constructions functioning as subject or com-
plement form absolute clauses. This is because this kind of word strings may have
two different interpretations: (1) The participial phrase functions as the post-mod-
ifier of the nominal group; and (2) the participial phrase functions as the logical
predicate of the nominal group. In the case of the former, the whole construction is
a nominal group, the noun being the head of the nominal group functioning as the
subject of the main verb, and the non-finite element being an embedded element
5.3  Dependent Absolute Clauses 77

functioning as post-modifier. For example, in 5-22a, the subject of the main verb
is Coracles, and the non-finite element using nets functions as post-modifier. In
the latter case, however, the noun and the non-finite element together function as
the subject of the main verb, forming an absolute clause, as Children having fun in
5-22b.
5-22a. Coracles using nets were banned from the Wye in the twenties. (BNC_NEWS)
b. Children having fun is quite a structured exercise. (BNC_MAG)
In SFL, the subject clause here is embedding, the whole construction being
a simple clause, but the complement clause is projection, the whole construc-
tion being a clause complex. In discussing the relationships of clause complexes,
Halliday and Matthiessen (2004: 443) define three projection systems: (i) the level
of projection (idea versus locution), (ii) the mode of projection (hypotactic report-
ing versus paratactic quoting), and (iii) the speech function (projected proposition
versus projected proposal). Thus, the system of projection is shown in Fig. 5.7.
5-23a. Ben said, “Why are you dressed like that?” (COCA_FIC)
b. Ben said he was working on the ninth of the twelve steps. (COCA_FIC)
c. Kemp said, “Don’t put ice in that.” (BNC_FIC)
d. I told you not to disturb me. (BNC_FIC)
e. Mller said, “Goodbye, Madame.” (BNC_FIC)
In 5-23a, the level of projection is locution, the mode of projection is paratactic
quoting, and the speech function is proposition (statement). In 5-23b, the level of
projection is idea, the mode of projection is hypotactic reporting, and the speech
function is proposition (statement). In 5-23c, the level of projection is locution, the
mode of projection is paratactic quoting, and the speech function is proposal. In
5-24d, the level of projection is idea, the mode of projection is hypotactic reporting,

locution
LEVEL
idea

reporting
MODE
projection
quoting
statement
proposition
major question
FUNCTION proposal
minor

Fig. 5.7  System of projection
78 5  Relationships Realized by Absolute Clauses

and the speech function is proposal. In addition to proposition and proposal, the
paratactic projected elements can also be minor speech functions.
Paratactic projected clauses can be used independently, without any signs
of being projected. In this sense, paratactic projected non-finite clauses have the
potential to form absolute clauses. However, non-finite clauses tend to realize
hypotaxis rather than parataxis and hence cannot form absolute clauses. The hypo-
tactic projected clauses of proposition are usually finite clauses, as in 5-24a. There
are only a small number of “accusative + infinitive” constructions that can be non-
finite clauses, as in 5-24b. The hypotactic projected clauses of proposal can be
finite ones, as in 5-24c, or non-finite ones, as in 5-24d.
5-24a. Martha thought that the river was angry. (BNC_FIC)
b. I believe it to be the source of evil. (BNC_FIC)
c. Lightman ordered that Kruger be publicly reprimanded on two counts.
(BNC_NEWS)
d. He told me to leave my job. (BNC_FIC)
As discussed in Chap. 2, traditional grammar defines absolute clauses in both
form and function. SFL considers language as meaning potential, emphasizing
using language to create meaning. It does not restrict the functions being able to
be realized by a language form. Both non-finite clauses in 5-24b and 5-24d have
their own subject. It is certain that not all non-finite clauses with subject can nec-
essarily form absolute clauses. The subject of the non-finite clause in 5-24b is it.
In traditional grammar, it is the object of believe, and to be the source of evil is the
complement of it. In SFL, the non-finite clause it to be the source of evil functions
as the complement of believe, and the non-finite clause itself has its own func-
tional syntactical structure.
I believe it to be the source of
evil
TG analysis Subject Predicate Object Complement
SFL analysis Subject Finite Main Complement
verb Subject Predicate
Transitivity Senser Process: mental Identified Process: relational Identifier
structure Metaphenomenon
Logico-semantic Primary Secondary
Interdependency α ‘β

It can be seen that traditional grammar considers 5-24b as a simple clause,


while SFL considers it as a hypotactic clause complex of projection. In this clause
complex, the subject it of the projected non-finite clause has no syntactic relation
with the main verb of the primary clause and hence can be regarded as an abso-
lute clause. To support this point, we attempt to carry on a transitivity analysis of
5-19d.
5.3  Dependent Absolute Clauses 79

Weather permitting, they shuttle up and down in


packs
Transitivity Circumstance Actor Process: Circumstance
structure Actor Process: material
material
Logico-semantic Secondary Primary
Interdependency ×β α

The non-finite clause in both 5-24b and 5-19d can be considered as an ele-
ment of the primary clause. Both have their own functional syntactic structure
and both are independent from the primary clause. The only difference is that the
non-finite clause in 5-24b realizes participant and that in 5-19d realizes circum-
stance. According to traditional grammar, 5-24b is a simple clause. Although the
object and complement form a subject–predicate construction, they cannot form
an absolute clause, because the object is controlled by the main verb. 5-19d is a
clause complex, and the subject of the logical subject–predicate construction is
not controlled by the main verb, and hence, they can form an absolute clause. The
secondary clauses of both clause complexes of expansion and of projection distin-
guished by SFL can be non-finite clauses. The adverbial clauses distinguished by
traditional grammar can be non-finite clauses, but the object clauses have no finite
equivalents. Because of this, absolute clauses defined by traditional grammar only
serve to function as adjuncts. According to SFL, the projected clauses functioning
as complement have both finite and nonfinite forms. Since absolute clauses can
realize circumstances, they should also be able to realize participants.
Weather permitting, they shuttle up and down
If weather permits, in packs
Transitivity Relator Actor Process: Actor Process: Circumstance
structure material material
Circumstance
Logico-semantic Secondary Primary
Interdependency ×β α
I believe it to be the source
Transitivity Senser Process: (that) it is of evil
structure mental Identified Process: Identifier
relational
Metaphenomenon
Logico-semantic Primary Secondary
Interdependency α ‘β

5-24d is a reported proposal. According to traditional grammar, it is a simple


clause like 5-24b. me is the object of the main verb, and to leave my job is the
complement of the object. According to SFL, 5-24d is a clause complex. Although
me is the logical subject of the nonfinite clause to leave my job, it is controlled by
the main verb told, and the nonfinite clause to leave my job is a projected clause,
and therefore, they cannot form an absolute clause.
80 5  Relationships Realized by Absolute Clauses

He told me to leave my
(that) I should leave job
Transitivity Sayer Process: Recipient Actor Process: Goal
structure verbal material
verbiage
Logico-semantic Primary Secondary
Interdependency α ‘β
The reported proposal can be locution or idea. For example, 5-24d is a reported
locution. The subject of a reported locution is implicit and it requires the recipi-
ent’s reasoning. For example,
5-25a. She had promised him to obey his commands. (COHA_FIC)
b. *He had been promised she to obey his commands.1
c. *He had been promised by her to obey his commands.
d. He had been promised that she would obey his commands.
5-26a. I promised him a wheelbarrow to be pushed every day in the resolution of
his debt. (COHA_FIC)
b. He was promised a wheelbarrow to be pushed every day in the resolution
of his debt.
Based on the above analysis, the recipient me in 5-24d is the subject of the pro-
jected non-finite clause, and the projecting clause can be changed to the passive
form I was told to leave my job. The subject I of the primary clause is still the
subject of the non-finite clause. The recipient him in 5-25 belongs to the primary
clause in structure, and it is not the subject of the non-finite clause, and therefore,
they do not form an absolute clause.
She had him to obey his
promised (that) would obey commands
she
Syntactic Subject Predicate Complement 1 Subject Predicate Complement
structure Complement 2
Transitivity Sayer Process: Recipient Actor Process: Goal
structure verbal material
Verbiage
Logico-semantic Primary Secondary
Interdependency α ‘β

The subject of a passive reporting locution is explicit, but the Actor is still implicit.
Since the subject of a passive clause is not constrained by the main verb of the project-
ing clause, passive non-finite clauses can form absolute clauses. See example 5-26:
I promised him a wheelbarrow to be pushed …
(that) a be pushed
wheelbarrow

1  * Unacceptable
5.3  Dependent Absolute Clauses 81

Syntactic Subject Predicate Complement 1 Subject Predicate


structure Complement 2
Transitivity Sayer Process: Recipient Goal Process: material
structure verbal verbiage
Logico-semantic Primary Secondary
Interdependency α ‘β

As can be seen, the Sayer she in 5-25a is the implicit subject of the projected
non-finite clause. If the projecting clause is changed into passive voice, the subject
of the primary clause is not the subject of the non-finite clause. In other words, the
subject of the non-finite clause is not co-referential with that of the primary clause.
Therefore, the subject of the non-finite clause should be made explicit. There are two
ways: First, changing the non-finite clause into an absolute clause; second, using
preposition by to introduce the explicit Agent. Following the first way, we get 5-25b.
According to Halliday and Matthiessen (2004), 5-25b is grammatically unac-
ceptable. This is because although she is the explicit subject of the non-finite
clause, it is also the Agent of promised. In a passive clause, the Agent and the
subject do not overlap, so the subject should not be in the nominative case. If the
subject of the non-finite clause is not the Agent of the main verb of the primary
clause, this construction is grammatically acceptable. For example,
5-27a. Wright was promised an opportunity to advance. (COHA_MAG)
b. ?To advance, Wright was promised an opportunity.2
5-28a. For some time we have been promised legislation to reform the laws on
friendly societies. (BNC_MISC)
b. To reform the laws on friendly societies, for some time we have been
promised legislation.
5-29a. The first 200 attendees were promised passes to the Tom Hanks film.
(COHA_NEWS)
b. *To the Tom Hanks film, the first 200 attendees were promised passes.
However, although 5-27a, 5-28a, and 5-29a are similar to 5-25b in form, the two
structures are different in syntactic function. The subject of the nonfinite clause to
advance in 5-27a is co-referential with that of the primary clause. However, syntacti-
cally to advance is embedded in the nominal group an opportunity functioning as its
post-modifier. The two elements together are projected by the main verb promised. In
other words, the projected element is not a macrophenomenon, but a phenomenon, and
the whole construction is a simple clause rather than a clause complex and hence can-
not form an absolute clause. This can be better demonstrated in 5-29a. 5-28a can also
be analyzed like this, except that the subject of the non-finite clause in 5-28a is not co-
referential with that of the primary clause. The subject of the non-finite clause to obey
his commands in 5-25b is not co-referential with that of the primary clause; to obey his
commands is the projected element of promised, and it is a macrophenomenon.

2  ? Maybe acceptable
82 5  Relationships Realized by Absolute Clauses

Following the second way, we get 5-25c. In 25c, the initiator her of the verb
promised is introduced by the preposition by, but by cannot at the same time intro-
duce the subject of the nonfinite clause. That is to say, the subject of the non-finite
clause remains implicit.
Therefore, the only way to make the subject of the non-finite clause explicit is
to change the non-finite clause into finite, as in 5-25d. From another perspective, if
the subject of the reported locution is explicit, it does not need the recipient’s rea-
soning. There will be a different situation. For example,
5-30 They promised her that the reign of terror would be over before the year
was out. (BNC_FIC)
5-31a. Russia was promised that NATO would not expand. (COCA_ACAD)
b. ?Russia was promised NATO not to expand.
In 5-26a, a wheelbarrow is the subject of the projected non-finite clause. It is
not directly controlled by the main verb. Therefore, the projected passive clause a
wheelbarrow to be pushed every day in the resolution of his debt may be consid-
ered as an absolute clause. The passive form of the projecting clause can still pro-
ject a non-finite clause, as in 5-26b and 5-32, because the subject of this projected
nonfinite clause is originally explicit.
5-32 He was offered counselling to be paid for by the church. (BNC_K1S)
Another interpretation of the infinitives in 5-27a and 5-28a is that they are expand-
ing clauses of purpose. For example, the two clauses in a hypotactic expanding
clause complex can be translocationed. This can be better explained in 5-33.
5-33a. He also promised measures to overcome two obstacles…. (BNC_NA)
b. He also promised measures were to be taken.
c. He also promised measures to be taken.
d. He also promised measures to overcome two obstacles were to be taken.
e. He also promised measures were to be taken to overcome two obstacles.
f. To overcome two obstacles, he also promised measures were to be taken.
In 5-33a, the string to overcome two obstacles functions as the post-modifier of the
projected element measures or as an expanding non-finite clause of purpose. In 5-33b,
the projected finite clause measures were to be taken can be changed into the non-finite
clause in 5-33c, forming an absolute clause. This is because measures and to be taken
form a non-finite clause with subject. It is obvious that to overcome two obstacles in
5-33d functions as post-modifier and that in 5-33e is an expanding non-finite clause of
purpose because it is not immediately after the head noun measures that it modifies, and
moreover, it can be translocated with the primary clause, as is shown in 5-33f.
We can give further explanations to absolute clauses of reported locution in
5-34 and 5-35.
5-34 The King held it good to accomplish her desire; and forthwith ordered let-
ters to be drawn up to Rodrigo of Bivar. (BNC_NA)
5-35 He ordered fighters to shoot down Hassans Boeing. (BNC_MISC)
5.3  Dependent Absolute Clauses 83

In 5-34, letters is not the object of ordered; rather, it is the subject of the non-
finite to be drawn up to Rodrigo of Bivar. They together form an absolute clause.
In 5-35, fighters is the object of ordered. It functions as the subject of the non-
finite clause to shoot down Hassans Boeing, but they do not form an absolute
clause because the non-finite clause to shoot down Hassans Boeing cannot be
translocated with the primary clause to realize the relationship of purpose.
Different from reported locution, the subject of a reported idea is explicit and
it is part of the projected proposal in structure. That is why the projecting clauses
cannot be changed into passive voice, but the projected clauses can. For example,
5-36a. He wanted me to design the top-floor atelier. (COCA_MAG)
b. He wanted the top-floor atelier to be designed by me.
According to traditional grammar, him in 5-25a, me in 5-36a, and the top-floor
atelier in 5-36b are all controlled by the main verb, they function as the object of
the main verb, and the infinitive to obey his commands in 5-25b and to be designed
by me in 5-36b both function as the complement of the object. According to SFL,
however, in 5-25a, him is the complement of the main verb and to obey his com-
mands is the projected clause, and me in 5-36a and the top-floor atelier in 5-36b
are included in the non-finite clauses but not controlled by the main verb in struc-
ture. Therefore, this kind of non-finite clauses with subject consisting of a nominal
group or a pronoun and a non-finite element can form absolute clauses, realizing
macrophenomena in transitivity structure.
He wanted me to design the top-floor atelier
Syntactic structure Subject Predicate Subject Predicate Complement
Complement
Transitivity Senser Process: Actor Process: Goal
structure mental material
Phenomenon
Logico-semantic Primary Secondary
Interdependency α ‘β

5.4 Embedded Absolute Clauses

Not only can two clauses constitute a clause complex realizing the logico-­semantic
relations of parataxis and hypotaxis, but also one of the two clauses can be rank-
shifted to embed in the other clause serving as a constituent of this clause or
embed in a constituent of the clause serving as the modifier of this constituent.
“There is no direct relationship between an embedded clause and the clause within
which it is embedded.” (Halliday and Matthiessen 2004: 426) Therefore, a clause
containing an embedded clause is still a simple clause but not a clause complex.
According to SFL, clauses are composed of groups or phrases, and a clause can
be rankshifted into a group to function as a constituent of another clause or can be
84 5  Relationships Realized by Absolute Clauses

further rankshifted into a word to function as a constituent of a constituent of the


clause. When a clause realizes the circumstance of another clause, the two clauses
form a clause complex; when a clause realizes a participant of another clause, it
functions as the subject or complement of this clause in structure, that is, nomi-
nal clauses in the traditional sense. Clauses realizing modifiers of nominal groups
are referred to as attributive clauses or adjectival clauses in traditional grammar.
Embedded clauses distinguished by Halliday and Matthiessen (2004) include
embedded bound clauses and embedded fact clauses. The former are embedded
in nominal groups to function as modifiers and the latter, in clauses to function as
participants which are actually the appositive of the antecedent “fact”.
5-37a. The people who had saved my life were my close relations! (BNC_FIC)
b. That the core of the issue was hedonism was unsurprising. (BNC_ACAD)
c. Nicandra regretted that her friend was not sleeping in the largest guest
room. (BNC_FIC)
d. The fact that women remarry is not surprising. (BNC_ACAD)
The modifier who had saved my life of the nominal group The people functioning
as the subject of the clause in 5-37a is an embedded bound clause. The embedded
clause in 5-37b does not function as a modifier; rather, it is the core element of the
nominal group, that is, it functions as the subject of the clause as a nominal group.
The embedded clause in 5-37c functions as the complement of the verbal group
regretted. The embedded clause in 5-37d is in fact the appositive of the core noun the
fact, which functions as the antecedent. In fact, the embedded clause in 5-37c is also
not directly projected by regretted; rather, it is the appositive of the omitted the fact.
Embedded clauses can also be non-finite. For example,
5-38a. Having a child is an enriching thing as regards that. (BNC_NA)
b. Cowley could see him sitting at the garden table. (BNC_FIC)
c. The animals coming into the infirmary were almost entirely horses.
(BNC_MISC)

5.4.1 Classification of Non-finite Verbs

The core element of a non-finite clause is the non-finite verb. Traditional gram-
mar identifies three types of non-finite verbs: participles, infinitives, and gerunds.
Participles can be further divided into present participles and past participles.
Present participles and gerunds both are the -ing form of verbs, and sometimes,
it is hard to distinguish between them. Traditional grammar also categorizes non-
finite verbs into -ing verbs, -ed verbs, and infinitives. The -ing verbs include pre-
sent participles and gerunds. The two ways of classification of non-finite verbs are
shown in Figs. 5.8 and 5.9.
The first classification tends to be function-oriented and the second, form-
oriented. Although the second classification has been universally accepted by
grammarians nowadays, the first classification is obviously briefer and clearer.
5.4  Embedded Absolute Clauses 85

Fig. 5.8  Classification of infinitive
non-finite verbs (I)
present
nonfinite verb participle
past

gerund

Fig. 5.9  Classification of infinitive
non-finite verbs (II)
present participle
nonfinite verb -ing
gerund

-ed

Infinitives and gerunds can both function as subjects or objects, and infinitives and
participles (present participles and past participles) can both function as adjuncts.
The syntactic functions of infinitives equal to the combination of those of gerunds
and participles.
SFL considers the traditional adjectival clauses and the nominal fact clauses as
embedded elements. As for non-finite clauses, SFL adopts the classification shown
in Fig. 5.9. It does not distinguish present participles and gerunds. For example,
Yang (2003: 52–54) distinguishes four types of constructions that non-finite -ing
and to- clauses as embedding can occur, they are subject, complement, modifier,
and the so-called anticipatory subject of cleft sentences. Non-finite -ed clauses
occur in two constructions, in which the non-finite -ed clause is embedded as mod-
ifier of the preceding element or as the complement of the clause. Although SFL
does not distinguish between gerunds and present participles, from these syntactic
functions, it can be seen that the non-finite participial (-ing and -ed) complement
functions as Attribute in relational clauses, while the gerundial complement real-
izes participant. For example,
5-39a. Not being fluent in English results in a language barrier that negatively
affects the patient’s educational process. (COCA_ACAD)
b. They don’t mind getting their hands dirty. (COCA_ACAD)
c. The door remains closed a minute and then there is a soft knock.
(COHA_FIC)
Non-finite elements in 5-39a and 5-39b act as subject and complement, respec-
tively, realizing participants. The former is the Actor of the material clause and the
latter, the Phenomenon of the mental clause. The non-finite element locked in 5-39c
functions as complement, realizing the Attribute of the relational verb remains.
The traditional appositive clauses “seem to have the hypotactic relation of elab-
oration, but the non-finite clauses are only used to describe groups in the finite
clauses” (Yang 2003: 53). Although appositive clauses and attributive clauses
86 5  Relationships Realized by Absolute Clauses

are the same in form, they are different in terms of rankshift. Appositive clauses
belong to the category of nominal clauses, the antecedents of which are usually
general nouns, such as the fact and the idea. Attributive clauses belong to the cat-
egory of adjectival clauses. According to SFL, appositive clauses realizing partici-
pants are at the group rank, while attributive clauses realizing modifiers are at the
word rank. As can be seen, although clauses realizing participants and those realiz-
ing modifiers are both embedded elements, they are different in embedding depth.
In the following section, we will distinguish types of embedded clauses from the
perspective of the embedding depth.

5.4.2 Types of Embedding

SFL considers clause complexes as composed of two interdependent clauses of the


same rank, while embedded clauses as an element of a simple clause. However,
there is not a clear dividing line between clause complexes and simple clauses.
Hypotactic clauses and embedded clauses are on a cline in rank scale. The basic
unit constituting a clause is group and the minimal unit is word. Clauses realizing
modifiers are embedded in nominal groups; they are at the word rank. Fact clauses
realizing participants are embedded in clauses; they are at the group rank. Clauses
realizing circumstances can also be considered being embedded in clauses; they
are at the rank of group/phrase. “A group is a bloated word, whereas a phrase is
a shrunken clause.” (Halliday 1994: 215), and phrases and groups are of the same
rank. Therefore, the dependent clauses in clause complexes can also be seen as
embedded elements of simple clauses. The traditional appositive clauses are at an
intermediate stage between the word rank (adjectival clauses) and the group rank
(nominal clauses). Functionally, they are in between hypotaxis and modifier, hence
“are actually a kind of loose embedding” (Yang 2003: 53). The syntactic relations
realized by embedded clauses are shown in Fig. 5.10.
A non-finite clause “may be rankshifted as a unit on the group layer, i.e.
embedded to a constituent; it may also function as the dependent clause in a clause
complex” (Yang 2003: 67). For example,
5-40 She doesn’t know what to do with her eyes. (COHA_FIC)
Thus, 5-40 can be seen as a clause complex or a simple clause. The non-
finite clause what to do is both projected and embedded. Thompson (1996): 23,
Fawcett (1996: 309), and Huang (1999: 121–123) all consider the dependency in
the Hallidayan sense as embedment, putting that the dependent clause in a clause
complex can also be regarded as a circumstance of the dominate clause. Biber
et al. (1999: 192) divide embedding into several levels, with hypotactic expansion
included. See example 5-41:
5-41a. Maya is drinking her first bourbon tonight because Vern left today for San
Francisco State. (Biber et al. 1999: 192)
b. That this was a tactical decision quickly became apparent. (ibid.: 193)
5.4  Embedded Absolute Clauses 87

clause Figure clause realizing patataxis

clause realizing circumstance


group/phrase Element
clause realizing participant

word Quality clause realizing modifier

Fig. 5.10  Syntactic relations realized by embedded clauses

c. They believe that the minimum wage could threaten their jobs. (ibid.: 193)
d. I have no idea when he will come back home. (ibid.: 193)
e. We have 30 men who are working from 6 am to 11 pm. (ibid.: 195)
f. And she said that everything was mouldy. (ibid.: 196)
It seems that the dependent clauses in clause complexes can all be taken as embed-
ded clauses, and therefore, hypotactic clause complexes can all be reanalyzed as sim-
ple clauses. The dependent clauses are rankshifted and embedded into the dominate
clauses. Only paratactic clauses cannot be seen as embedded elements. The syntactic
functions of the secondary clauses in 5-41 are shown in the following figures.
5.41a Maya is drinking her bourbon tonight because Vern left today
Syntactic Subject Predicate Complement Adjunct Subject Predicate Adjunct
structure Adjunct
Transitivity Actor Process: Goal Circumstance Relator Actor Process: Circumstance
structure material material
Circumstance

5-41b That this was a tactical quickly became apparent


decision
Syntactic Subject Predicate Complement Adjunct Predicate Complement
structure Subject
Transitivity Relator Carrier Process: Attribute Circumstance Process: Attribute
structure relational relational
Carrier

5-41c They believe that the minimum wage could threaten their jobs
Syntactic Subject Predicate Subject Predicate Complement
structure Complement
Transitivity Senser Process: Relator Actor Process: material Goal
structure mental Phenomenon

5-41d I have no idea when he come back home


Syntactic Subject Predicate Antecedent Subject Predicate Adjunct
structure Appositive
Complement
88 5  Relationships Realized by Absolute Clauses

Transitivity Possessor Process: Relator Actor Process: Circumstance


structure relational material
Possessed

5-41e We have 30 men who are working from 6 am to


11 pm
Syntactic Subject Predicate Antecedent Subject Predicate Adjunct
structure Modifier
Complement
Transitivity Possessor Process: Actor Process: Circumstance
structure relational material
Possessed

5-41f And she said that everything was mouldy


Syntactic Subject Predicate Subject Predicate Complement
structure Complement
Transitivity Sayer Process: Relator Carrier Process: Attribute
structure verbal relational
Verbiage

5-41a is a clause complex of hypotactic expansion. The dependent clause real-


izes circumstance in the dominate clause. 5-41c and 5-41f are clause complexes
of projection projecting an idea and a locution, respectively. However, both 5-41c
and 5-41f are nominal clauses (object clauses). If a construction with an embed-
ded complement clause can be considered as a clause complex, as in 5-41c and
5-41f, then a construction with an embedded subject clause can also be considered
as a clause complex, as in 5-41b. According to Biber et al. (1999), 5-41e is two-
level embedding, and the embedded clause is not a direct element of the primary
clause; rather, it is embedded in an element of the primary clause to function as
the modifier of the element within which it is embedded. The embedded clause in
5-41d functions as appositive. Although an appositive clause belongs to the cat-
egory of nominal clause, it can still be considered as two-level embedding. This is
because the appositive clause is not a direct element of the primary clause; rather,
it describes, supplements, or explains an element of the primary clause.
He (2002) refers to the construction composed of an embedded clause and a
clause within which it is embedded in the Hallidayan sense as a clause complex.
Seen from this point, the traditional adverbial clauses, subject clauses, and object
clauses can all be double-analyzed in transitivity and syntax. That is, they can all
be considered as clause complexes or simple clauses with embedded elements.
Attributive clauses and appositive clauses in traditional grammar are not direct
elements of the clauses within which they are embedded. Therefore, constructions
with attributive clauses or appositive clauses are still simple clauses. This is also
true for non-finite clauses. For example,
5-42a. I can’t imagine anything happening to any of them. (COCA_FIC)
b. To know the right is not enough. (COCA_FIC)
c. I do not mind wearing my English clothes. (COCA_FIC)
d. He seemed to enjoy the exercise, playing with his daughter. (COCA_MAG)
e. Mr. Fedders had borrowed heavily since joining the S.E.C. (COCA_NEWS)
5.4  Embedded Absolute Clauses 89

5-42a I can’t imagine anything happening to any of them


Syntactic structure Subject Finite Predicate Subject Predicate Complement
Complement
Transitivity Senser Process: mental Actor Process: Circumstance
structure material
Phenomenon

5-42b To know the right is not enough


Syntactic structure Predicate Complement Predicate Complement
Subject
Transitivity structure Process: mental Phenomenon Process: relational Attribute
Carrier

5-42c I don’t mind wearing my English


clothes
Syntactic structure Subject Finite Predicate Predicate Complement
Complement
Transitivity structure Senser Process: mental Process: material Goal
Phenomenon

5-42d He seemed to enjoy the exercise, playing with his


daughter
Syntactic Subject Finite Predicate Antecedent Predicate Adjunct
structure Appositive
Complement
Transitivity Senser Process: mental Process: Circumstance
structure material
Phenomenon

5-42e Mr. Fedders had borrowed heavily since joining the S.E.C.
Syntactic Subject Finite Predicate Complement Predicate Complement
structure Adjunct
Transitivity Carrier Process: relational Attribute Relator Process: Goal
structure material
Circumstance

Absolute clauses in traditional grammar can be considered as the embedded ele-


ments realizing circumstance in SFL. Embedded non-finite verbal groups with
subject realizing participants can form absolute clauses. An embedded element
realizing modifier or appositive is not the direct constituent of the clause within
which it is embedded. If the embedded element is a non-finite clause with subject
which is not the element within which the embedded element is embedded, i.e.,
the antecedent, this embedded clause has the potential to form an absolute clause.
However, this is not necessarily true. This is because when a clause, no matter
whether it is non-finite or finite, functions as the post-modifier of a noun, this noun
as the antecedent of this embedded clause always plays some certain role in this
embedded clause, such as subject, complement, or adjunct.
90 5  Relationships Realized by Absolute Clauses

5-43a. Birds which are sold into the pet trade are notoriously badly treated.
(BNC_NA)
b. The book which he wrote was circulated to the English universities.
(BNC_MISC)
c. The days when he could do it, or would, were now over. (BNC_FIC)
The antecedents Birds, the book, and the days in 5-43 function as subject, com-
plement, and adjunct, respectively. When the antecedent functions as the subject of
the embedded clause, the embedded clause can be non-finite, as in 5-44, but when
the antecedent functions as complement or adjunct, there are no non-finite forms.
Therefore, non-finite clauses functioning as post-modifier cannot form absolute
clauses.
5-44 People standing nearby laughed softly in derision. (COCA_FIC)
Appositive clauses are relatively special. On the one hand, they belong to the
category of nominal clauses, and on the other hand, they are not the direct partici-
pants of the main verbs. They are embedded in the nominal groups functioning as
participants of the clauses. For example,
5-45a. I thought that he was only a captain. (COCA_FIC)
b. We all regretted that he was left behind. (COCA_FIC)
c. I regretted the fact that I was late. (COCA_FIC)
The project clause in 5-45a is an idea and that in 5-45b is a fact. According
to the foregoing analysis, the former is one-level embedding and the latter, two-
level embedding. It is not directly projected by the projecting clause Mark Antony
regretted; rather, it is the embedded element of the phenomenon the fact, as in
5-45c. The difference between appositive clauses and modifier clauses in form
lies in that the antecedent of the former does not assume an element of the clause,
while that in the latter, assumes an element. Therefore, appositive non-finite
clauses with subject have the potential to form absolute clauses. For example,
5-46 ?The fact, Caesar being ambitious, worries me so much.
Appositive clauses are between modifier clauses and complement clauses in
rank. Modifier clauses have no structural relations with the clauses in which the
nominal groups within which they are embedded assume functions. Subject and
complement clauses both assume syntactic functions in the clauses within which
they are embedded. Appositive clauses are intermediate between them two; they
may assume syntactic functions or not. For example,
5-47a. I can’t enjoy your favorite exercise, swimming in winter. (Zhang 1997: 1355)
b. He had one aim—to fight for the independence of his motherland. (ibid)
In 5-47a, swimming in winter is the appositive of your favorite exercise. It
can also realize participants, that is, they two have the same syntactic function
potential. In 5-47b, to fight for the independence of his motherland is the apposi-
tive of one aim, but they do not realize participants, that is, they are syntactically
5.4  Embedded Absolute Clauses 91

different. However, both appositives function to explain an element of the main


clause, that is, your favorite exercise and one aim, respectively. As is discussed in
Sect. 5.3.1.1, absolute clauses of elaboration may function to explain the primary
clause as a whole (see example 5-4b) or just one or more of its constituents (see
example 5-6). When an absolute clause functions to explain a constituent of the
primary clause, it is actually an absolute clause of appositive.
The difference between appositive clauses and modifier clauses in the possi-
bility to form absolute clauses is that the former have the potential to form abso-
lute clauses, but the latter has not. The difference between appositive clauses and
complement clauses is that the former are facts and the constructions of nominal
group + non-finite element can certainly form absolute clauses, and the latter are
objective phenomena and the constructions of nominal group + non-finite element
do not necessarily form absolute clauses.
5-48 She had always considered Richard to be a fine and good man. (BNC_FIC)
The nominal group + non-finite element construction Richard to be a fine and
good man in 5-48 is an embedded non-finite complement clause. It is a fact clause
and hence forms an absolute clause.
Subject clauses and complement clauses are both subordinate clauses in tra-
ditional grammar. However, SFL considers subject clauses as embedded clauses,
which are at the group rank, acting as an element in the clauses within which they
are embedded, while considers complement clauses as projected clauses, which are
at the clause rank, acting as the hypotactic clauses in clause complexes. In fact,
both the subject and the complement are participants in transitivity and they should
be at the same rank. If subject clauses are accepted as embedded clauses, then there
is no reason to exclude complement clauses from embedded clauses. If changed
into passive clauses, the original complement becomes the subject in the passive
clauses. If the construction containing a complement were a clause complex, then
the passive clause with a subject clause could also be referred to as a clause com-
plex. This is to say that it is appropriate to consider the projected clauses as embed-
ded clauses. If this interpretation is acceptable, then the hypotactic clauses in the
Hallidayan sense of clause complexes can also be seen as embedded clauses.
It would appear that in all the relationships between clauses, only parataxis can-
not be interpreted as embedding. This is because neither of the two clauses can be
considered as an element of the other. The criterion to identify embedding is dou-
ble transitivity analyses. That is one of the two clauses can be an element of the
other clause or an element of an element of the other clause. All hypotactic clauses
can be double-analyzed in transitivity structure, so can the embedded modifier or
appositive clauses. It is only paratactic clauses that cannot be double-analyzed.
However, this is not necessarily true. Paratactic projected clauses seem to be able
to be interpreted as one element of the projecting clauses, i.e., the complement.
Therefore, paratactic projected clauses can be double-analyzed in transitivity. For
example,
92 5  Relationships Realized by Absolute Clauses

5-49a. Then she said, “I’ll see you in five weeks.” (BNC_FIC)
b. It was stood close beside her in the grass, and said, “Hello, Peggie.”
(BNC_MISC)
However, paratactic projection is different from paratactic expansion at least in
the following four aspects. First, the two clauses in a paratactic clause complex of
expansion can be linked by the conjunction and, but the two clauses in a paratactic
clause complex of projection cannot. Second, both the two clauses in a paratactic
clause complex of expansion can be independently used, but the primary clause in
a paratactic clause complex of projection cannot. Third, the two clauses in a para-
tactic clause complex of expansion cannot be transpositioned, and even if they can,
the one in front position is always the primary clause, but the two clauses in a para-
tactic clause complex of projection can be transpositioned, and the status of each of
the two clauses will not change. Fourth, the two clauses in a paratactic clause com-
plex of expansion are at the same rank, but in a paratactic clause complex of projec-
tion, the projected clause can be a finite one as well as other constructions, such as
minor clauses. However, minor clauses are not clauses in the strict sense, and they
are at the same rank as the projecting clauses. Therefore, the parataxis of projec-
tion is actually not real parataxis. They share many features of hypotaxis and hence
belong to a special parataxis between parataxis and hypotaxis in the real sense.
From the above analysis, parataxis of projection can also be considered as
embedding, but not deep embedding. From the perspective of embedding depth,
all embedding relations can form a cline (see Fig. 5.11).
Embedded clauses defined by SFL include embedded defining clauses, apposi-
tive clauses, and subject clauses, but its definition on complement clauses is rela-
tively unclear. For example, Halliday and Matthiessen (2004) refer to complement
clauses as projected clauses, but some projected non-finite clauses as embed-
ded clauses. In fact, Fig. 5.11 shows that the dependency of absolute clauses is
changing gradually with that of the depth of embedding; there are no absolutely
independent absolute clauses. These above seven types of structure are shown sys-
tematically in Fig. 5.12.
When one of the two clause in the “clause + clause” structure is non-finite, this
relation system will greatly change, as shown in Fig. 5.13.
As can be seen, the rank status of non-finite clauses is lower than that of finite
clauses. This is mainly shown in that non-finite clauses cannot realize paratac-
tic expansion, projected non-finite clauses are characterized with hypotaxis,
and hypotactic non-finite clauses of projection are treated as simple clauses by
Halliday and Matthiessen (2004), etc.
From the above analysis, we find that except for the clause complexes of para-
tactic expansion, other relation types of clause complexes and simple clauses with
clausal subject, appositive, and modifier can be regarded as embedding. They
embed in different depths. In these types of embedded clauses, only the embed-
ded clauses of modifier cannot form absolute clauses because the antecedents of
the embedded clauses always assume certain syntactic functions in the embedded
clauses. Other types of embedded clauses can all form absolute clauses.
5.4  Embedded Absolute Clauses 93

Fig. 5.11  Cline of Clause + Paratactic clause of expansion


embedding depth
Paratactic clause of projection

Adjunct clause

Complement clause

Subject clause

Appositive clause

Modifier clause

Fig. 5.12  Relation system of expansion


“clause + clause” structures parataxis
projection

adjunct
clause +clause hypotaxis
complement

subject
embedding appositive

modifier

Fig. 5.13  Relation system of ----


“clause + non-finite clause” parataxis
structures
projection

adjunct
clause + nonfiniteclause hypotaxis
complement

subject

embedding appositive
modifier
94 5  Relationships Realized by Absolute Clauses

5.5 Continuing Absolute Clauses

Logico-semantic relation is a semantic concept. Since functional syntax does not


limit the function types of absolute clauses, then except for hypotaxis and embed-
ding, can absolute clauses realize parataxis?
Clauses can be used independently, can constitute clause complexes with
another clause to realize parataxis and hypotaxis, and can also be rankshifted to
embed in another clause to function as an element of that clause or to embed in
a constituent of that clause to function as the modifier of this constituent. Two
independent clauses do not constitute a clause complex, but they are not always
clustered together randomly; they too realize certain interdependent and logico-
semantic relations. For example,
5-50a. I will wear an evening dress. However, I don’t like the long dress.
(BNC_NEWS)
b. At first I thought it was steam from the coffee machine. Then I realised it
must be a ghost. (BNC_NEWS)
The two independent clauses in 5-50a realize parataxis. The cohesive relation
between the two clauses is adversative which is realized by the cohesive adverb
however. 5-50b consists of two clause complexes, and the cohesive relation
between them is additive. Both 5-50a and 5-50b realize parataxis of extension.
Absolute clauses as non-finite clauses cannot realize parataxis. However, in a
clause complex, there are two linking devices, one being the co-referential sub-
ject and the other, the conjunctive expression. Even if there is not an explicit con-
junctive expression between the two clauses, they can still only realize hypotaxis.
There is only one linking device, that is, the conjunctive preposition, between a
non-finite clause with subject and the primary clause. If there is not an explicit
conjunctive preposition, the non-finite clause with subject can stand alone.
Therefore, in a clause complex with an absolute clause as the secondary clause,
both the primary clause and the secondary clause tend to be independent and
hence are likely to realize parataxis.
“In hypotaxis, the two clauses, primary and secondary, can occur in either
order: either α ^ β or β ^ α.” (Halliday 1994: 220; Halliday and Matthiessen 2004:
379) “In parataxis, only the order 1 ^ 2 is possible” (ibid), and the secondary (con-
tinuing) clauses elaborate, extend, or enhance the primary (initiating) clauses. In
paratactic of projection, the two clauses can occur in either order: either 1 ^ 2 or
2 ^ 1.
5-51a. John ran away, which surprised everyone. (Halliday 1994: 222)
b. John ran away, and this surprised everyone.
According to SFL, the interdependency between the two clauses is hypotaxis,
and the logico-semantic relation is elaboration. The dependent clause functions as
the description of the dominant clause. Huang (1998c) however believes that the
interdependent relation between the two clauses in 5-51a is parataxis, the primary
5.5  Continuing Absolute Clauses 95

clause being the initiating clause and the secondary clause being the continuing
clause, and the logico-semantic relation between the two clauses is not elaboration
but extension. There are two reasons. First, the two clauses cannot be translocated.
For example, 5-51a can be rewritten as 5-51b. Second, in a hypotactic clause com-
plex, only the primary (dominant) clause is free, and the secondary (dependent)
clause cannot occur independently. However, clauses like which surprised every-
one are able to be used independently.
5-52a. 
But to buy some more you need more cash. Which you haven’t got.
(BNC_FIC)
b. 
It is therefore upward rankshift; which we must consider. (Halliday
1966[2002]: 121)
Even if the logico-semantic relationship between the two clauses is not exten-
sion but elaboration, the clause introduced by which can still only be considered
realizing parataxis. For example,
5-53a. He was at the zoo, which had just been given a rare species of bird.
(COCA_FIC)
b. *He was at the zoo, and it had just been given a rare species of bird.
The two clauses in 5-53a cannot be translocated and hence is paratactic in
Huang’s (1998a, b, c) terms, and the continuing clause which had just been given
a rare species of bird functions as the explanation of the zoo, so the logico-seman-
tic relation between the two clauses is elaboration. 5-53b is unacceptable because
the conjunction and realizes extension. In this sense, the hypotactic clause com-
plexes consisting of two finite clauses discussed by Halliday (1994: 227–229) and
Halliday and Matthiessen (2004: 400–403) actually realize parataxis, and the log-
ico-semantic relation is elaboration or extension (Huang 1998a, b, c).
Instances 5-1d and 5-53a in traditional grammar are both nonrestrictive attribu-
tive clauses. The difference between the two is that the antecedent of the second-
ary clause of the former is the whole primary clause; it adds information to the
primary clause, realizing extension. The antecedent of the secondary clause of the
latter is a nominal clause in the primary clause; it is the explanation of the nominal
clause, realizing elaboration. The absolute clause the gerbils running in 5-6 and
those in 5-54 cannot be translocated with the primary clauses and hence are abso-
lute clauses of elaboration. This can also be used to explain why these absolute
clauses cannot be introduced by with. Then, can these absolute clauses be consid-
ered realizing parataxis?
5-54a. She could hear voices ahead, a girl squealing, a man laughing.
(BNC_FIC)
b. The pose is natural—relaxed—the legs slightly apart, the arms hanging
down. (COCA_FIC)
These absolute clauses are all relational clauses with Subject and Finite o­ mitted,
and the residues the gerbils running, a girl squealing, and a man laughing, etc.,
96 5  Relationships Realized by Absolute Clauses

themselves form absolute clauses. At the same time, absolute clauses can be used
independently. For example,
5-55 There was a lot of background noise. People chatting, laughing. (BNC_FIC)
Although the absolute clause in 5-55 elaborates the preceding clause semantically,
the two clauses do not form a clause complex. They are two independent clauses
realizing parataxis. The question is can the translocatability and independability
be the necessary and sufficient conditions for the two clauses to realize parataxis?
The reason why the which clauses can be considered continuing is that in addition
to meeting the two conditions, they should be finite clauses, which is an impor-
tant precondition. Absolute clauses are non-finite. They are at a lower level in rank
than finite clauses. Therefore, whether absolute clauses can realize parataxis also
requires to be supported by whether non-finite clauses and finite clauses can be
paratactic. In the following, we will try to analyze the functional syntactic struc-
ture of 5-56.
5-56 Frank sat reading the newspaper. (Quirk et al. 1985: 1126)
According to traditional grammar, 5-56 is a simple clause. The predicate sat
cannot introduce an object because it is intransitive, and the present participle
phrase reading the newspaper functions as an attendant circumstance, illustrating
the action state of the event sat. However, according to SFL, there are two differ-
ent processes in 5-56. One is realized by a finite element and the other by a non-
finite element. “So it is a clause complex when embedding is excluded. Here the
relator and the process are conflated and they are realized together by reading”
(Yang 2003: 59). However, Yang (2003) has not discussed the relation between
the two clauses in this clause complex. In Halliday’s thinking, a non-finite clause
can only realize hypotaxis in a clause complex. The non-finite clause in 5-56 real-
izes additive extension. In hypotaxis, two clauses can be translocated, but the two
clauses in 5-56 cannot. This is because if translocated, the non-finite clause read-
ing the newspaper will realize enhancement. The logico-semantic relationship will
not change when a dependent clause is translocated with the dominant clause. This
proves that the clause complex consisting of a finite clause and a non-finite clause
in 5-56 realizes paratactic extension.
Parataxis is different from hypotaxis in that the former “is logically (i) sym-
metrical and (ii) transitive” (Halliday and Matthiessen 2014: 452). When we say
the two clauses in a paratactic clause complex cannot be translocated, we mean the
initiating clause always occurs first. Absolute clauses are always secondary. If an
absolute clause can occur in front of the primary clause, it is certainly dependent.
It should be noted that the distinction of relation types of clause complexes in
SFL is meaning-based and that in traditional grammar is form-based. Seen from
the formal perspective, “subordinate clauses are finite, non-finite, or verbless”
(Greenbaum 1996: 306), but “co-ordination and subordination can sometimes
express similar meaning relationships” (ibid). The relationship between the finite
clause and the non-finite clause in 5-56 is subordination in form, while parataxis in
meaning.
5.5  Continuing Absolute Clauses 97

Thus, clause complexes consisting of a finite clause and an absolute clause


can realize both parataxis and hypotaxis. The interdependency realized by abso-
lute clauses can be tested with the preposition with. According to SFL, with is a
conjunctive preposition introducing non-finite clauses. The non-finite clauses with
subject introduced by with can only realize hypotaxis. Absolute clauses of exten-
sion and enhancement can be translocated with the primary clauses, hence are
hypotactic, and can be introduced by with. Absolute clauses of elaboration cannot
be translocated with the primary clauses, hence are paratactic, and cannot be intro-
duced by with. For example,
5-57 He had to use other men’s property, the use being right. (COHA_NF)
5-58 Six other people followed suit, the women crying out, the men groaning.
(COHA_FIC)
The absolute clause in 5-57 functions as an explanation of the primary clause
and that in 5-58 as an explanation of the nominal group in the primary clause.
These absolute clauses and those in 5-6 and 5-54 are appositive clauses. This
means that the interdependent relation of absolute clauses of elaboration is actu-
ally not hypotactic, but paratactic. Although appositive clauses are deep embedded
(see Fig. 5.11), appositive absolute clauses (absolute clauses of elaboration) have
a strong tendency to be independent. In the next section, we will analyze the inde-
pendence of different relation types of absolute clauses.

5.6 Independence of Absolute Clauses

From the above analysis, it is clearly seen that the independence of absolute
clauses changes with the depth of embedding. There are no absolutely independent
absolute clauses. Finite clauses are in principle independent and have a tendency
to be used alone, while non-finite clauses are dependent in principle. Absolute
clauses are non-finite clauses. The so-called absolute is relative to the primary
clause to which an absolute clause is attached. In the three types of absolute con-
structions distinguished by traditional grammar, the subject of a free adjunct con-
struction and that of the primary clause can be co-referential, hence a related free
adjunct construction, or not be co-referential, hence an unrelated free adjunct con-
struction, or dangling construction. It is obvious that the two types of free adjunct
constructions are not the same in independence.
Augmented absolute constructions are not the absolute constructions in the
strict sense because they are introduced by the conjunctive preposition with.
Absolute clauses are not introduced by conjunctive expressions and their sub-
jects are not co-referential with those of the primary clauses. Therefore, absolute
clauses are more independent compared with the other two types of absolute con-
structions. However, absolute clauses of different syntactic functions are differ-
ent in independence. In this section, we will discuss the independence of absolute
clauses and their tendency to be independent.
98 5  Relationships Realized by Absolute Clauses

5.6.1 Dimensions of Independence of Absolute Clauses

Absolute clauses have a multi-function potential. In the interdependent relations,


absolute clauses can realize both hypotaxis and parataxis. In the logico-semantic
relations, absolute clauses can realize both expansion and projection. In simple
clauses, absolute clauses can realize both circumstance and participant. At the text
level, absolute clauses can be used independently and realize some certain logico-
semantic relations with the preceding or following clauses through textual cohe-
sive devices. Therefore, according to the syntactic functions realized by absolute
clauses, we can distinguish four types of absolute clauses: absolute adjunct clauses
(including absolute clauses of hypotactic extension and absolute clauses of hypo-
tactic enhancement), absolute appositive clauses (absolute clauses of elaboration),
absolute complement clauses (absolute clauses of projection), and absolute subject
clauses (absolute clauses of embedding) (see Fig. 5.14).
These types of absolute clauses are different in independence. The independ-
ence of absolute clauses is embodied in two dimensions: (1) the requirement of a
subject and (2) the requirement of a conjunctive expression.

5.6.1.1 Dimension of Subject

The independence of the hypotactic clauses in 5-59 increases with the decreasing
requirement of the subject.
5-59a. Since I left my hospital position, I have made my living by writing.
(COCA_MAG)
b. When questioned, Ray evaded and changed the subject. (COCA_FIC)
c. On reaching Bongi Street, the vehicle was hit violently by a light truck.
(COCA_ACAD)
d. Judging from the way he rode the train, the guy knew martial arts.
(COCA_FIC)

Free adjunct

Absolute appositive clause

Absolute adjunct clause


Absolute construction Absolute clause
Absolute subject clause

Absolute complement clause

Augmented absolute

Fig. 5.14  System of types of absolute construction


5.6  Independence of Absolute Clauses 99

The hypotactic clause in 5-59a is finite, and the subject cannot be omitted. The
conjunction when in 5-59b realizes relator, and the logical subject of the hypo-
tactic non-finite clause is the nominative case and is co-referential with that of
the primary clause. The preposition on in 5-59c realizes relator, and the omitted
subject may be the oblique case, as in on him reaching Bongi Street, or posses-
sive, as in on his reaching Bongi Street. Although “the preferred form in current
usage is the ‘oblique’ case” (Halliday and Matthiessen 2004: 421), the requirement
of a subject of this kind of non-finite clauses has clearly decreased. The subject
of a dependent non-finite clause is usually co-referential with that of the primary
clause and hence is typically omitted. The non-finite clause in 5-59d has not a con-
junctive expression, and its subject is not co-referential with that of the primary
clause. The omitted subject may be we or a general pronoun one. Crystal (2008: 2)
refers to this kind of constructions independent from the primary clauses as abso-
lute constructions, but not absolute clauses for lacking an explicit subject.

5.6.1.2 Dimension of Relator

The independence of the hypotactic clauses in 5-60 increases with the decreasing
requirement of the conjunctive expression.
5-60a. Since I’ve come back, I haven’t worked a day. (COCA_NEWS).
b. Because of bylaws requiring a window, he decided to leave an open,
32,000 square foot inner core. (COCA_NEWS)
c. With money being so tight, you have to shop around. (COCA_NEWS)
d. Compromise having failed, there was left only force. (BNC_ACAD)
The finite secondary clause in 5-60a has the strongest requirement of a rela-
tor. If there is not an explicit conjunction, the relation between the two clauses is
not hypotactic, but paratactic. Despite that the non-finite clause in 5-60b has its
own subject, the relator is realized by the preposition because of which makes the
dependent clause not being able to be independent from the dominate clause in
structure. If the preposition is removed, the non-finite clause becomes an absolute
clause. In 5-60c, as in 5-60b, the non-finite clause with subject is introduced by a
preposition with, but with “does nothing to narrow down the range of logical roles
which it may assume” (Stump 1985: 13). It only realizes relator which prevents
the non-finite clauses with subject from forming absolute clauses. The non-finite
clause with subject in 5-6d is not introduced by an explicit conjunctive expression
and is independent from the primary clause in structure, hence forming an abso-
lute clause. The logico-semantic relation can be inferred from the meaning of the
primary clause. For example,
5-61a. John went off by himself, the rest of us staying behind. (Halliday and
Matthiessen 2004: 404)
b. Lu found an empty one and got in, her mother following. (COCA_FIC)
100 5  Relationships Realized by Absolute Clauses

The primary clause in 5-61a emphasizes that John leaves by himself, and the abso-
lute clause functions as the explanation of this proposition. The logico-semantic
relation is elaboration. The absolute clause in 5-61b supplements the information
of the primary clause, and the logico-semantic relation is extension.
Absolute clauses can form clause complexes with finite clauses realizing elab-
oration, extension, and enhancement. The difference between an absolute clause
and a subjectless non-finite clause is that the former is more independent. It may
not be introduced by a conjunctive expression and can be used independently with
all the basic characteristics of parataxis. In Chap. 6, we will carry out a qualitative
and quantitative research of the relationships realized by absolute clauses.
The dependence of dependent clauses increases along two lines starting from
finite clauses. One is the decreasing of the requirement of a subject and the other
is the decreasing of the requirement of a conjunctive expression. The first line
ends at absolute non-finite clauses, i.e., dangling constructions in traditional gram-
mar, and the second line ends at absolute clauses. Both belong to the category of
absolute constructions. The difference between the two is that absolute clauses
are semantically not attached to any element of the primary clause and are not
intended to modify any nominal group. An absolute non-finite clause, however, is
purposed to modify a certain nominal group, but is wrongly linked up with another
nominal group.

5.6.2 Independent Tendency of Absolute Clauses

Different function types of absolute clauses we distinguished according to the cri-


teria of identifying absolute clauses worked out in Chap. 3 are different in inde-
pendence. The independent tendency of absolute clauses is also embodied in two
dimensions: (1) the dimension of absolute clauses and (2) the dimension of pri-
mary clauses.

5.6.2.1 Dimension of Absolute Clauses

In clauses composed of nominal groups, verbal groups and adverbial groups, the
adverbial groups realizing circumstances are relatively loosely connected with
the verbal groups realizing processes, and their positions are relatively flexible. A
clause without circumstances is still structurally integrated. In clause complexes
with absolute clauses realizing circumstances, the absolute clauses have no formal
links with the primary clauses and they are not indispensible in meaning either.
Therefore, absolute clauses of circumstance are strong in independence. They are
usually separated from the primary clauses with a comma in form and are also
flexible in position: An absolute clause can be preceding or following the primary
clause, or even in between the primary clauses. Without the absolute clause, the
primary clause still enacts a complete proposition or proposal.
5.6  Independence of Absolute Clauses 101

Absolute clauses of circumstance realize the logico-semantic relation of exten-


sion and enhancement in clause complexes. Absolute clauses realizing these two
relations are different in independence. For example,
5-62a. Trade having been done, they settled to take the meal. (BNC_FIC)
b. She was seated, upright in her chair, eyes wide open. (BNC_MISC)
The absolute clauses in 5-62a and 5-62b realize enhancement and extension,
respectively. The former is a relation of time and the latter, attendant circumstance.
According to the criteria of identifying absolute clauses, the finite equivalent of
the former is a dependent clause of time and the latter, a continuing clause of
extension (see example 5-63).
5-63a. After trade was done, they settled to take the meal.
b. She was seated, upright in her chair, and her eyes were wide-open.
Two clauses of parataxis are equal in status and cannot be translocated. Two
clauses of hypotaxis are unequal in status and can be translocated, but the log-
ico-semantic relation will not change. Two clauses of parataxis can both be free
from each other, while of the two clauses of hypotaxis, only the primary clause is
free and the secondary clause is dependent. From this sense, when the continuing
clause and the dependent clause change into absolute clauses, the logico-semantic
relation will not change although the interdependent relation changes. Thus, abso-
lute clauses of extension are more independent than those of enhancement.
Absolute clauses realizing participants are also different in independence. For
example,
5-64a. I heard the wind rustling in the bushes. (BNC_MISC)
b. The Christian considers discipleship to be a vocation. (BNC_MISC).
5-64a and 5-64b are both clause complexes of projection. The former is a mac-
rophenomenon and the latter, a metaphenomenon. In 5-64a, the wind has double
functions. First, it functions as the subject of the non-finite clause, enacting a prop-
osition. And at the same time, it functions as the complement of the main verb.
Since traditional grammar considers the non-finite clause as the complement of the
object and the complement is not obligatory, then I heard the wind is syntactically
acceptable. From the perspective of SFL, the water functions as the subject of rus-
tling in the bushes. They form a subject + predicate construction which as a whole
functions as the complement of heard. Although I heard the wind is syntactically
acceptable, the meaning is different because what we heard is the sound of water
but not the water itself. In this sense, the wind rustling in the bushes can be con-
sidered as an absolute clause. Nevertheless, the independence of this kind of abso-
lute clauses is relatively weak.
In 5-64b, the two considers in considers discipleship and considers discipleship
to be a vocation are different in meaning. 5-64b is a clause complex of projection,
the projected clause being an idea. Compared with 5-64a, 5-64b is more independ-
ent. Since discipleship is not the complement of considers, then why cannot the
nominal group discipleship following considers omitted? There are two reasons.
102 5  Relationships Realized by Absolute Clauses

One is that the subject of the non-finite clause is not co-referential with that of
considers, so it requires its explicit subject. Even if the subject of the non-finite
clause is co-referential with that of considers, the non-finite clause still requires
an explicit subject, the slot of which can only be filled with the reflexive pronoun
himself. The other is that what the verb considers projects is a fact, but not an act.
However, infinitives are perfective and hence cannot construe a fact. It is only
when they have their own subject that they can construe a complete proposition of
fact. For example,
5-65a. Sidney considers trying hard to be good spirited. (COCA_FIC)
b. *Sidney considers to try hard to be good spirited.
The subject of the non-finite clause and that of the main verb of the project-
ing clause are co-referential in 5-65. The gerund in 5-65a is imperfective and
can construe a fact proposition, functioning as the complement of the main verb
considers. The infinitive in 5-65b is a projected act, and so it cannot construe a
fact proposition, nor function as the complement of the main verb considers. This
further indicates that although a project act can also form an absolute clause, the
independence of a projected fact is stronger.
The relationship between the subject of non-finite clauses and the main verb of
the primary clauses is a vital factor to affect the independence of absolute clauses.
5-66a. He wanted me to be a lawyer. (COCA_FIC)
b. He invited me to stay for a few days. (COCA_MAG)
c. To stay for a few days, he invited me.
The word me in 5-66a is less closely tied with the main verb in the project-
ing clause than that in 5-66b. The former first of all functions as the subject of
the infinitive, and the construction as a whole functions as the complement of the
main verb in the projecting clause and hence is more independent. The latter first
of all functions as the complement of the main verb in the projecting clause and
the infinitive to stay for a few days can be considered as the purpose of the main
verb, realizing enhancement. The whole structure can be changed into 5-66c.
As can be seen from 5-64 and 5-66, projected clauses with subject are possible
to form absolute clauses, the independence of which is determined by the relation
of the subject of the non-finite clause and the main verb. The relation between the
subject of the non-finite clause in 5-66b and the main verb is the weakest, and
therefore, the independence of the non-finite clause with subject is the strongest.
The subject of the non-finite clause in 5-66a belongs only to the non-finite clause,
but this non-finite clause with subject construes an act rather than a fact and the
independence is relatively weaker. The subject of the non-finite clause in 5-64a is
also the complement of the main verb heard, but this complement is not always
necessary. For example, changing 5-64a into The wind rustling in the bushes I
heard is acceptable, while changing it into Rustling in the bushes I heard the wind
is unacceptable. From this point, the non-finite clause in 5-64a is still an abso-
lute clause, except that the independence is much weaker. The non-finite clause
in 5-66b is also a subject + predicate construction, but it cannot form an absolute
5.6  Independence of Absolute Clauses 103

clause because me cannot be separated from the main verb. Thus, the non-finite
clauses with subject in the four example sentences are on a cline in independence:
5-64b > 5-66a > 5-64a > 5-66b
Seen from the process types realized by verbs, non-finite clauses projected by ver-
bal verbs cannot form absolute clauses and those projected by mental verbs can
form absolute clauses. This fits well with the independence rules of the projected
phenomenon: Verbal verbs project acts, mental verbs project facts, and projected
facts are more independent than projected acts.
Absolute clauses functioning as subject and appositive are both fact clauses.
Both are more independent than absolute clauses functioning as complement. As
for subject and appositive, the former is more independent than the latter. The rea-
son is that although the two are both embedded clauses, the subject of the latter
is still controlled by the main verb to some extent. For example, if discipleship
in 5-64b is a personal pronoun, it should be accusative case. The subject of the
former is not controlled by the main verb; it can be nominative case and hence is
more independent.
According to this analysis, we can make this conclusion: Seen from the inde-
pendent tendency, different relation types of absolute clauses form a cline:
Elaboration > Extension > Enhancement > Embedding > Projection (Fact > Act)

5.6.2.2 Dimension of Primary Clauses

In constructions containing two clauses, the two clauses are interdependent. When
one of the two clauses is an absolute clause, its independence differs in terms of
the different relations between clauses. The independence of absolute clauses is
also embodied in that of the primary clauses.
In constructions containing absolute clauses, the absolute clauses can realize
circumstances and participants as well. “The configuration of process + partici-
pants constitutes the experiential centre of the clause.” (Halliday and Matthiessen
2004: 176) Circumstances are peripheral and they are not involved in the pro-
cesses. The interdependence between adverbial groups realizing circumstances
and verbal groups realizing processes is relatively weak. The absence of the cir-
cumstance will not affect the integrity of the clausal structure. That is to say, in
constructions containing absolute clauses realizing circumstances, the primary
clauses are less dependent on the absolute clauses. Similarly, in a clause com-
plex of parataxis, each of the two clauses has its own syntactic structure, so dou-
ble transitivity analyses is not allowed. Since absolute clauses of extension can
be transformed into continuing clauses, in clause complexes containing absolute
clauses of extension, the primary clauses is the least dependent on the absolute
clauses and hence is the most independent. In clause complexes with absolute
clauses of enhancement, the absolute clauses can be transformed into finite
clauses, realizing the circumstance of the main verb, and the whole construction
can be double-analyzed in transitivity. Removing the absolute clauses will affect
104 5  Relationships Realized by Absolute Clauses

the basic configuration of the primary clauses, but the structure is still complete.
Therefore, the primary clauses are still independent of the absolute clauses.
Compared to circumstances, participants are relatively closer to processes in
relation. In the unmarked case, participants are closely bound to the processes
and hence are indispensible. From the perspective of the relationship between the
subject and the main verb, absolute clauses functioning as subject are more inde-
pendent of the primary clauses than those functioning as complement, but from
the requirement of participants by the main verbs, the constructions with abso-
lute clauses functioning as subject are less independent than those with absolute
clauses functioning as complement. Here is the reason. Although the subject and
complement of a verbal group are both indispensible, in the same syntactic struc-
ture, the same constituents can be omitted. This is to say, both the subject and the
complement can be omitted. However, the omission of the subject is structural,
which can only be realized in paratactic clause complexes, while the omission of
the complement may be cohesive. This shows that a verb requires the subject more
intensely than the complement, and hence, a clause without complement is more
independent than that without subject. When the omitted elements are obvious, a
clause without complement can be used alone, but a clause without subject cannot.
For example,
5-67a. He stopped and took off his watch. (COCA_FIC)
b. Millions of people watch the event (Sun., March 21, 8 p.m., ABC) on TV
around the world. Who will win and who deserves to win? (COCA_NEWS)
5-67a is a paratactic clause complex. The subject of the second clause is co-
referential with that of the first clause and hence is omitted. 5-67b consists of two
separate sentences, the second sentence of which is a paratactic clause complex.
The complement the event in both of the two clauses in this clause complex is
omitted. Ellipsis is a cohesive device at the textual level.
Appositive clauses are not direct participants of the main verbs, and so their
relationship with the main verbs is relatively loose. In constructions with absolute
clauses of appositive, the absolute clauses are strongly independent. The clauses
within which they are embedded are also strongly independent. The independence
of different kind of primary clauses is also on a cline:
Elaboration > Extension > Enhancement > Projection > Embedding
Despite the fact that the subject of an absolute subject clause has nothing to do
with the main verb, this kind of absolute clause is still strongly independent.
However, the intense requirement of a subject by the main verb weakens the inde-
pendence of the construction as a whole. It is obvious that of the two types of
independent tendencies, the primary clause-based independent tendency is more in
line with the characteristics of absolute clauses.
5.7 Summary 105

5.7 Summary

In this chapter, we analyzed syntactic and functional types of absolute clauses


according to the meaning potential of absolute clauses as non-finite clauses from
such perspectives as absolute clauses as dependent clauses, absolute clauses as
embedded clauses, and absolute clauses as continuing clauses. The functional
analysis of absolute clauses can be used to effectively answer the questions in the
research of absolute clauses in tradition grammar. For example, the interdependent
relation between absolute clauses of elaboration and the primary clauses is para-
tactic, hence able to be introduced by with. Absolute clauses of additive and adver-
sative and those of enhancement can be introduced by with. with does not change
the grammatical status of absolute clauses as hypotactic clauses or the logico-
semantic relation with the primary clauses. The meaning of with is only reflected
in the independence of the secondary clauses from the primary clauses. Absolute
clauses as non-finite clauses can be seen as embedded elements in all construc-
tions. They can not only realize circumstances but also realize participants,
functioning as appositive, adjunct, subject, and complement, forming absolute
appositive clauses, absolute adjunct clauses, absolute subject clauses, and absolute
complement clauses. There are no absolute post-modifier clauses.

References

Biber, D., Johansson, S., Leech, G., Conrad, S., & Finegan, E. (1999). Longman grammar of spo-
ken and written English. London: Longman.
Crystal, D. (2008). A dictionary of linguistics and phonetics (6th ed.). Oxford: Blackwell
Publishing.
Curme, G. O. (1931). A grammar of the English language: Syntax. Boston: D. C. Heath.
Fawcett, R. P. (1996). A systemic functional approach to complementation in English. In
M. Berry, C. Butler, R. P. Fawcett & G. W. Huang (Eds.). Meaning and form: Systemic func-
tional interpretations (pp. 297–366). Norwood, NJ.: Ablex.
Greenbaum, S. (1996). The Oxford English grammar. Oxford: Oxford University Press.
Halliday, M. A. K. (1966[2002]). The concept of rank: A reply. In J. Webster (Ed.), Collected
works of M. A. K. Halliday, vol. 1: On grammar (pp. 118–126). London: Continuum.
Halliday, M. A. K. (1994). An introduction to functional grammar (2nd ed.). London: Edward
Arnold.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2004). An introduction to functional grammar
(3rd ed.). London: Edward Arnold.
Halliday, M. A. K., & Matthiessen, C. M. I. M. (2014). Halliday’s introduction to functional
grammar (4th edn.). London and New York: Routledge.
He, W. (2002). On Recursiveness, Embedding and Functional Syntactic Reanalysis. Foreign
Language Research, 3, 64–69.
Huang, G.-W. (1998a). Form is the realization of meaning. Foreign Languages and Their
Teaching, 9, 4–7.
Huang, G.-W. (1998b). A functional analysis of the English causative structure. Journal of
Foreign Languages, 1, 12–16.
Huang, G.-W. (1998c). A functional analysis of the “wh—continuing clause” in English. Modern
Foreign Languages, 1, 1–9.
106 5  Relationships Realized by Absolute Clauses

Huang, G.-W. (1999). Explorations in English language and linguistics. Guangzhou: Zhongshan
University Press.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London and New York: Longman.
Reuland, E. J. (1983). Governing -ing. Linguistic Inquiry, 14, 101–136.
Stump, G. T. (1985). The semantic variability of absolute constructions. Dordrecht: D. Reidel
Publishing Company.
Thompson, G. (1996). Introducing functional grammar. London: Edward Arnold.
Yang, B.-J. (2003). A study of non-finite clauses in English: A systemic functional approach.
Beijing: Foreign Language Teaching and Research Press.
Zhang, Z.-B. (1997). A new English grammar (3rd ed.). Shanghai: Shanghai Foreign Language
Education Press.
Chapter 6
Absolute Clauses Distributed
in Three Corpora

6.1 Introduction

In Chap. 2, we reviewed the study of absolute clauses in traditional grammar in


terms of syntactic functions, stylistic distribution, historical distribution and case
choice, etc. There are still problems with these studies. For example, traditional
grammar fails to reach an agreement on the case of absolute clauses and it does
not account for the role of with in augmented absolute clauses. Since absolute
clauses are subordinate constructions with no finite verb, why do they merely act
as clausal adjuncts and circumstances?
In Chap. 3, on the basis of functional syntactic theory, we discussed the condi-
tions for forming absolute clauses, defined absolute clauses from the perspective
of SFL, and proposed three criteria of identifying absolute clauses. According to
the meaning potential of absolute clauses, we in Chap. 5 discussed the functional
syntactic types of absolute clauses within the framework of functional syntactic
theory and recognized four structural types of absolute clauses: absolute adjunct
clauses, absolute subject clauses, absolute complement clauses, and absolute
appositive clauses. Absolute clauses in the traditional sense cover such two types
as the absolute adjunct clauses and absolute appositive clauses, which are hypo-
tactic clauses of expansion, with absolute adjunct clauses realizing extension and
enhancement, and absolute appositive clauses realizing elaboration. Absolute com-
plement clauses are in fact some of those projected non-finite clauses.
In Chap. 5, we distinguished three types of projected absolute clauses: (1)
­proposition: idea, (2) proposal: locution, and (3) proposal: idea. In this chapter,
we will conduct a corpus-based quantitative research on the absolute clauses of
elaboration, extension, enhancement, and projection to illustrate their stylistic
­
distribution, historical distribution, and case choice based on the Brown Family
Corpora, the BNC and the COHA.

© Springer-Verlag Berlin Heidelberg 2015 107


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_6
108 6  Absolute Clauses Distributed in Three Corpora

6.2 Research Based on Brown Family Corpora

Using regular expressions 2–6 we wrote in Chap. 4, we have retrieved 382 abso-


lute clauses of elaboration, extension, and enhancement from the Brown Family
Corpora. However, it should be noted that, due to the limitation of these regular
expressions, there are some absolute clauses excluded out of the 382 instances.
Here are some examples:
6-1a. The boy blushed, a red tide rolling up his neck and into his ears. (FROWN _P)
b.  Catherine having been cast out, Georgina reigned in her stead undisputed
queen of the home, the children, and all official social affairs (LOB_G).
c. Fixing their bayonets as they ran the Devons made the frontal attack, a thun-
derstorm shielding them somewhat from the hail of rifle-fire (FLOB_E).
d. The half-man lay on the ground, arteries pulsating prettily behind the half-
transparent film which protected his abdominal cavity (BROWN_M).
e. There would be an ordered improvement in standards for households in all
income groups, each household moving to a house a little better than the
one it previously lived in (LOB_J).
As for the absolute clauses of projection, we will collect only the collocations
of a noun or pronoun, a projection verb and a personal pronoun followed by an
­infinitive, present participle, or past participle.
In the projection types, quoted clauses cannot form absolute clauses. The dif-
ference between reported ideas and reported locutions lies in that ideas are macro-
phenomena or metaphenomena projected by mental verbs and locutions verbiage
by verbal verbs. A mental verb cannot project at the same time a phenomenon and
a macrophenomena or metaphenomena. For example, the verb think can only pro-
ject a macrophenomena while believe can project a phenomenon or a macrophe-
nomenon. However, a verbal verb can have a recipient and a verbiage at the same
time. A projected non-finite macrophenomenon or metaphenomenon can form an
absolute clause, but a projected non-finite verbiage cannot because the recipient
remains unchanged (see the following examples).
6-2a. She thinks that we may have sent her too many book tokens (BNC_NA).
b. He told me that he hates dancing! (BNC_FIC)
As is discussed in Chap. 5, no matter whether they are propositions or propos-
als, reported locutions do not tend to form absolute clauses. Only can some pas-
sive non-finite clauses form absolute clauses, see examples 6-3 and 6-4:
6-3a. As I gave my gun to the officials to keep, I warned them to be careful with
it (BNC_FIC).
b. Hartlepool magistrates ordered him to carry out 240 h community service
and pay 250 compensation (BNC_NEWS).
6.2  Research Based on Brown Family Corpora 109

6-4a. When the emperor saw a flagstone engraved with a cross he ordered it to be
raised (BNC_MISC).
b.  When the king commanded deer to be taken for his use in the royal
­forests it was usually the warden’s responsibility to see that this was done in
the proper manner (BNC_NA).
Therefore, of the three types of absolute clauses of projection, we only retrieve
those of reported ideas, including proposition and proposal. There might be too
many concordance lines retrieved, if all the verbal groups projecting ideas are con-
sidered, so we just select several most representative verbs that can project both
finite clauses and absolute clauses, such as believe, consider, expect, feel, find,
hear, know, see, want. Then, we write the following regular expression:
RE7.\S+_(PP|N\w+)\s\S+_V\w*\s\S+_PP\s(\S+_TO\s\S+_V[BDHV]|\S
+_V(B|D|H|V)[GN])
Using this regular expression, we have collected 533 absolute clauses of projec-
tion from the Brown Family Corpora. In the next subsections, we will observe the
distributions of these absolute clauses.

6.2.1 Overall Distributions

In this section, we will analyze the overall distributions of the 382 absolute clauses
of expansion and the 533 absolute clauses of projection extracted from the Brown
Family Corpora, including the formal distribution, the historical distribution, and
the stylistic distribution.

6.2.1.1 Formal Distribution of Absolute Clauses of Expansion

By means of the regular expressions, we can extract all absolute clauses of expan-
sion with noun or pronoun subject and present participle, past participle, infinitive,
adjective, adverb, preposition phrase or noun predicate. The frequencies of all the
forms of absolute clauses are shown in Table 6.1.
Table  6.1 shows that, out of the 382 absolute clauses of expansion, there are
224 with non-finite verb predicate and 158 with verbless predicate, respectively.
The present participle stands out with 152 instances, accounting for 39.8 % of
the total, with the past participle, preposition phrase, adjective, and adverb fol-
lowing in sequence. The statistics also shows that absolute clauses with infini-
tive and noun predicate are the smallest in number. All the forms except infinitive
can be considered as absolute clauses with the present participle being omitted.
Therefore, some grammarians also refer to absolute clauses in the traditional sense
as absolute participles (see Zandvoort 1972).
The absolute clauses of projection can be classified in form by the projection
verbs selected (see Table 6.2).
110 6  Absolute Clauses Distributed in Three Corpora

Table 6.1  Formal Form Frequency Percentage (%)


distribution of absolute
Non-finite verb Present 152 224 39.8 58.6
clauses of expansion
participle
Past participle 69 18
Infinitive 3 0.8
Verbless Noun 3 158 0.8 41.4
Preposition 66 17.3
phrase
Adjective 58 15.2
Adverb 31 8.1
Total 382 100

Table 6.2  Formal Form Frequency


distribution of absolute
Believe 15
clauses of projection
Consider 6
Expect 37
Feel 16
Find 145
Hear 31
Know 8
See 42
Want 233
Total 533

Table 6.3  Historical and 1961 1991/1992 2009 Total


regional distribution of
Expansion American 69 62 69 200
absolute clauses
British 58 55 69 182
Sub-total 127 117 138 382
Projection American 96 75 108 279
British 87 89 78 254
Sub-total 183 164 186 533
Total 310 281 324 915

6.2.1.2 Historical Distribution

The historical and regional distributions of the 382 absolute clauses of expansion
and the 533 absolute clauses of projection are shown in Table 6.3.
Table 6.3 shows that absolute clauses of expansion are evenly distributed in the
six corpora. Seen from the regional distribution, British English is slightly out-
numbered by American English in 1961 and 1991/1992, while in 2009, they are
6.2  Research Based on Brown Family Corpora 111

equal in number. Seen from the historical perspective, from 1961 to 1991/1992,
both American English and British English are somewhat on the decline, but from
1991/1992 to 2009, both go steadily up with American English ending up increas-
ing to the level of 1961 and British English exceeding the level of 1961. Although
the total number is increasing in the span of nearly 50 years, it does not show any
significant regular tendency in historical and regional distributions.
Absolute clauses of projection are also evenly distributed in both region and
time. British English is slightly outnumbered by American English in 1961
and 1991/1992, while in 2009, American English dominates British English.
Historically, from 1961 to 1991/1992, American English is somewhat on the
decline, but from 1991/1992 to 2009, it goes up over the level of 1961, and British
English is right opposite. In general, neither absolute clauses of expansion nor
those of projection show significant difference in either regional or historical
distributions.

6.2.1.3 Stylistic Distribution

Table 6.4 aims to interpret the distribution of the absolute clauses in such four reg-
isters as press, general prose, learned and fiction.
Table  6.4 shows that the stylistic distribution of absolute clauses varies
greatly, mostly appearing in fiction, with general prose, press and learned fol-
lowing. However, there is little difference between American English and British
English in all the four styles. For example, except that British English outnum-
bers American English by one instance of both expansion and projection and seven
instances of projection in press, in the other three styles in expansion, American
English slightly outnumbers British English, and in the other two styles in projec-
tion, British English outnumbers American English. In general, the stylistic dis-
tribution of absolute clauses shows no significant difference between American
English and British English or between expansion and projection.
In fact, the original frequencies of absolute clauses in the four registers can
hardly bring an adequate explanation that absolute clauses are scarcely used in
press or learned, because in the six corpora, there are relatively fewer words in the
press and learned sub-corpora and as such, the number of the absolute clauses in

Table 6.4  Stylistic distribution of absolute clauses in the Brown Family Corpora


Press General prose Learned Fiction Total
Expansion American 14 43 12 131 200
British 15 38 8 121 182
Sub-total 29 81 20 252 382
Projection American 27 79 7 166 279
British 28 50 3 173 254
Sub-total 55 129 10 339 533
Total 84 210 30 591 915
112 6  Absolute Clauses Distributed in Three Corpora

250

200

150

100

50

0
Press General Prose Learned Fiction

Expansion 26.76 32.05 20.44 161.75


Projection 50.75 51.05 10.22 217.59

Fig. 6.1  Stylistic distribution of absolute clauses (per million words)

these two registers is surely smaller. Therefore, to facilitate effective comparison,


we convert the original frequencies of absolute clauses in the four styles into the
standard frequencies of per million words, as is shown in Fig. 6.1.
It is observed from Fig. 6.1 and Table 6.4 that although the absolute number
of absolute clauses in general prose is much larger than that in press and that in
learned, when converted into the standard frequency of per million words, the
difference among the three is significantly narrowed. Nevertheless, fiction is still
dominant in number. Seen from Fig. 6.1, the two lines are nearly in parallel, with
the learned being an exception. On the whole, absolute clauses show significant
differences in stylistic distribution. They do not tend to be used in learned, espe-
cially those of projection.

6.2.2 Functional Distribution

In this section, we will analyze the functional distribution of the 915 absolute
clauses extracted from the Brown Family Corpora, including the regional distribu-
tion, historical distribution and stylistic distribution, etc.

6.2.2.1 Regional Distribution of Functions

In Chap. 2, we categorized the functions of absolute clauses in the traditional


sense into attendant circumstances, clausal adjuncts, and appositives. According to
the SFL criteria of identifying absolute clauses we discussed in Chap. 3, we iden-
tified absolute clauses of expansion, projection, and embedding. It is not easy to
collect absolute clauses of embedding from the corpus, so in this section, we will
6.2  Research Based on Brown Family Corpora 113

Table 6.5  Regional distribution of functions of absolute clauses


Expansion Projection Total
Enhancement Extension Elaboration Sub-total
American 19 122 59 200 279 479
British 34 115 33 182 254 436
Total 53 237 92 382 533 915

analyze the distributions of absolute clauses of expansion and projection extracted


from the Brown Family Corpora (see Table 6.5).
Table  6.5 shows that absolute clauses of both expansion and projection are
nearly evenly distributed in region, with American English slightly outnumber-
ing British English. Of the three types of expansion, extension counts the most
and then comes elaboration, and enhancement is the least. However, as there is
no elaboration distinguished in traditional grammar, absolute clauses of elabora-
tion are included into those of extension, which indirectly explains why there are
relatively fewer absolute clauses of adverbials of time, condition, and cause. This
is possibly because of the influence of ablative absolutes in Latin. Ablative abso-
lutes are always translated into with phrases in English, which are recognized by
Stump (1985) as augmented absolute constructions, realizing the logico-semantic
relation of extension in terms of SFL. Seen from the perspective of regional dis-
tribution, the number of absolute clauses of enhancement in American English is
obviously smaller than that in British English, while the number of extension and
that of elaboration are definitely the opposite. Nevertheless, extension dominates
in both American English and British English.

6.2.2.2 Historical Distribution of Functions

In this section, we will analyze the historical distributions of the function types of
absolute clauses in the Brown Family Corpora (see Table 6.6 and Fig. 6.2).
Table 6.6 and Fig. 6.2 show that absolute clauses of enhancement, elaboration,
and projection nearly keep steady in number, and only those of extension show
a trend to increase. The number of elaboration in 1991/1992 is slightly larger
than that in 1961 but undergoes a decrease in 2009. The number of extension in
1991/1992 is a bit smaller than that in 1961 but undergoes a substantial increase
in 2009. Enhancement is on a gradual decline throughout the nearly 50 years.
Although the overall frequency of absolute clauses manifests no distinct changes
along time, yet from the perspective of function types, the total number of the
absolute clauses of extension is on the increase and that of enhancement is on the
decrease.
Figures 6.3 and 6.4 show that the four functions of absolute clauses share the
similar distribution trend along time in British English and American English. The
number of absolute clauses of enhancement in American English is on a steady
decrease, and that of extension in British English keeps increasing but that in
114 6  Absolute Clauses Distributed in Three Corpora

Table 6.6  Historical distribution of functions of absolute clauses


1961 1991/1992 2009 Total
LOB Brown FLOB Frown CLOB Crown
Expansion Enhancement 12 9 11 7 11 3 53
Extension 29 44 32 31 54 47 237
Elaboration 17 16 12 24 4 19 92
Sub-total 58 69 55 62 69 69 382
Projection 87 96 89 75 78 108 533
Total 145 165 144 137 147 177 915

200
180
160
140
120
100
80
60
40
20
0
1961 1991/1992 2009
Enhancement 21 18 14
Extension 73 63 101
Elaboration 33 36 23
Projection 183 164 186

Fig. 6.2  Historical distribution of functions of absolute clauses

American English goes at first downward and then upward. An irregular distri-
bution trend is shown in projection. In British English, it goes downward, and in
American English, first downward and then upward.

6.2.2.3 Stylistic Distribution of Functions

In this section, we will analyze the stylistic distribution of the four function types
of absolute clauses.
For a good visual presentation, we change the total numbers of enhancement,
extension, and elaboration into 533, the total number of projection (see Fig. 6.5).
Table 6.7 and Fig. 6.5 show that these four function types of absolute clauses
differ greatly in stylistic distribution. All function types of absolute clauses tend
6.2  Research Based on Brown Family Corpora 115

100
90
80
70
60
50
40
30
20
10
0
1961 1991/1992 2009
Enhancement 12 11 11
Extension 29 32 54
Elaboration 17 12 4
Projection 87 89 78

Fig. 6.3  Historical distribution of functions of absolute clauses in British English

120

100

80

60

40

20

0
1961 1991/1992 2009
Enhancement 9 7 3
Extension 44 31 47
Elaboration 16 24 19
Projection 96 75 108

Fig. 6.4  Historical distribution of functions of absolute clauses in American English

to be used in fiction, with extension dominating the other three and projection
following, and none of the four functions tends to occur in learned, especially
extension and projection. This is right opposite in fiction. Of the four functions,
the number of extension is the smallest in press, general prose, and learned. In
press and learned, enhancement counts the most and in general prose, elaboration.
The number of elaboration is the largest in general prose and the second largest in
learned. This is because fiction is characterized with narration of events and depic-
tion of actions and hence advisable to use absolute clauses of extension, while
learned is characterized with argumentation and hence favorable to use absolute
116 6  Absolute Clauses Distributed in Three Corpora

450
400
350
300
250
200
150
100
50
0
Press General Prose Learned Fiction
Enhancement 91 151 70 221
Extension 31 72 9 421
Elaboration 35 197 52 249
Projection 55 129 10 339

Fig. 6.5  Stylistic distribution of functions of absolute clauses

Table 6.7  Stylistic distribution of functions of absolute clauses


Press General prose Learned Fiction Total
Expansion Enhancement 9 15 7 22 53
Extension 14 32 4 187 237
Elaboration 6 34 9 43 92
Sub-total 29 81 20 252 382
Projection 55 129 10 339 533
Total 84 210 30 591 915

clauses of enhancement and elaboration. The stylistic distribution of functions of


absolute clauses will be further discussed in the following sections with more data
extracted from larger corpora such as the BNC and the COHA.

6.2.3 Personal Pronoun Subject and Its Case Choice

Traditional grammar fails to reach a final agreement on the case of the personal
pronoun subject of absolute clauses. It is popularly accepted that the subject of an
absolute clause is a zero noun or a nominative pronoun, or occasionally an accu-
sative pronoun. This at least shows that it is uncommon for accusative pronouns
to function as the subject of absolute clauses. In this section, we will analyze
the distribution of the personal pronoun subject of absolute clauses and its case
choice, including the historical distribution and the stylistic distribution. The case
of the personal pronoun subjects of absolute clauses of projection is the assigned
6.2  Research Based on Brown Family Corpora 117

accusative by the projection verbs, so we count only the personal pronoun subjects
of absolute clauses of expansion which are not assigned case by the main verbs in
the matrix clauses.

6.2.3.1 Distribution of the Personal Pronoun Subject

Historical Distribution

Among the 382 absolute clauses of expansion extracted from the Brown Family
Corpora, there are merely 13 instances with personal pronoun subject. This is not
enough to conduct statistical analyses. Therefore, it is necessary to appropriately
relax the restrictions on the retrieval and to rewrite the following regular expres-
sion by which all the collocations consisting of a personal pronoun and a non-
finite or verbless element can be retrieved.
RE8.\S + _(,|SENT|:)\s\S + _PP(\s\S + _RB\w*)*\s(\S + _TO\s\S + _V[BDHV]|\S 
+ _(V(B|D|H|V)[GN]|JJ\w*|IN|DT|N\w +))
Using this regular expression, we extracted 61 absolute clauses with personal
p­ronoun subject from the Brown Family Corpora, including 31 in American
English and 30 in British English. Table 6.8 shows the historical and regional dis-
tributions of the 61 absolute clauses. The number of finite clauses with personal
pronoun subject extracted from the Brown Family Corpora reaches 143,568. The
ratio of absolute clauses with personal pronoun subject and the finite clauses is
4.25 to 10,000. This indicates that personal pronouns are not commonly used as
the subject of English absolute clauses. This is mainly because personal pronouns,
as anaphora in the precedent clauses, usually refer to a nominal group which is
usually a participant of the main verbs in the primary clauses. In the four types of
probable subjects distinguished by Kortmann (1991: 92), the co-referential sub-
jects of absolute clauses and the primary clauses are mostly personal pronouns,
and absolute clauses with personal pronoun subject mostly realize elaboration, that
is, absolute appositive clauses.
Table 6.8 shows that the overall distribution of absolute clauses with ­personal
pronoun subject is relatively uniform. There are only two more instances in
1991/1992 than in 1961, and in 2009, the total number remains the same. In addi-
tion, the number in American English is larger than that in British English in both
1961 and 1991/1992, and even is on a slightly increase in 1991/1992, but in 2009,
it undergoes a clear decrease, which is outnumbered by British English. Although
the number in British English remains the same in 1961 and in 1991, it is smaller

Table 6.8  Historical 1961 1991/1992 2009 Total


distribution of personal
American English 10 12 9 31
pronoun subjects of absolute
clauses British English 9 9 12 30
Total 19 21 21 61
118 6  Absolute Clauses Distributed in Three Corpora

than that in American English, and in 2009, it increases and outnumbers American
English. Nevertheless, on the whole, the distribution manifests no distinct regional
or historical changes.

Stylistic Distribution of Personal Pronoun Subjects

The distribution of the 61 personal pronoun subjects in the four registers of press,
general prose, learned, and fiction is shown in Table 6.9.
Table  6.9 shows significant register differences of absolute clauses with per-
sonal pronoun subject. Such absolute clauses are mostly used in fiction and gen-
eral prose. In fiction, American English outnumbers British English, but in press
and general prose, British English takes the lead in number. In learned, abso-
lute clauses with personal pronoun subject are inactive, with only two inanimate
third person reflexive pronouns in American English and none in British English.
Undoubtedly, these original data cannot fully explain the fact that absolute clauses
with personal pronoun subject are ill-performed in press and learned, because the
vocabulary of the two registers in the Brown Family Corpora is smaller, and so
the absolute number of personal pronouns is also smaller. Therefore, to facilitate
effective comparison, the original frequency will also be converted into the stand-
ard frequency of per million words, as is shown in Fig. 6.6.

Table 6.9  Stylistic distribution of personal pronoun subjects of absolute clauses


Press General prose Learned Fiction Total
American English 2 7 2 20 31
British English 3 15 0 12 30
Total 5 22 2 32 61

30

20

10

0
Learned Press General prose Fiction Average

American English 4.17 3.79 5.66 26.46 10.03


British English 0 5.68 12.14 15.87 10
Average 2.09 4.74 8.9 21.12 10.02

Fig. 6.6  Stylistic distribution of personal pronoun subjects of absolute clauses (per million words)
6.2  Research Based on Brown Family Corpora 119

Figure 6.6 shows that American English shares the same frequency with British
English in terms of personal pronoun subjects of absolute clauses. The frequency
reaches the highest in fiction which outnumbers the total of the other three styles.
In fiction and learned, American English exceeds British English in number, while
in press and general prose, American English falls behind. In American English,
only fiction stands over the average level and gains the dominant advantage
against the other three styles which are evenly distributed nearly to form a straight
horizontal line shown in the figure. Meanwhile, in British English, the four styles
of learned, press, general prose, and fiction present an arithmetic increase trend in
sequence, with general prose and fiction standing over the average level, which
implies that in comparison with American English, British English tends to use
absolute clauses with personal pronoun subject in particular in general prose.

6.2.3.2 Distribution of the Case of Personal Pronouns

Regional Distribution

Among the 61 instances of the absolute clauses with personal pronoun subject,
except 12 unmarked cases such as absolutive you and it as well as their corre-
sponding reflexives, there are 49 instances of case-marked first person and third
person pronouns, including 28 nominatives, 12 accusatives, and 9 reflexives, as is
shown in Table 6.10.
Table 6.10 shows that the case of personal pronoun subjects of absolute clauses
shares similar trend in both American English and British English. Nominatives
count the most, then accusatives follow, and reflexives stand the least. In addition,
the case mark of reflexives is usually accusative, because in finite clauses, reflex-
ives can alone be used as object rather than subject. Therefore, of the 49 case-
marked personal pronouns, there are actually 28 nominatives and 21 accusatives,
as is shown in Table 6.11.

Table 6.10  Regional Nominative Accusative Reflexive Total


distribution of the case
American 13 6 4 23
of personal pronoun subjects
English
of absolute clauses
British English 15 6 5 26
Total 28 12 9 49

Table 6.11  Regional Nominative Accusative Total


distribution of nominative
American English 13 10 23
and accusative personal
pronoun subjects of absolute British English 15 11 26
clauses Total 28 21 49
120 6  Absolute Clauses Distributed in Three Corpora

Table  6.11 shows that of the 49 case-marked personal pronouns, American


English and British English account for 23 and 26, respectively, and nominatives and
accusatives are evenly distributed, with the numbers of both nominative and accusa-
tive pronouns in British English slightly larger than those in American English.

Historical Distribution

The historical distribution of the case of personal pronoun subjects of absolute


clauses in the Brown Family Corpora is shown in Table 6.12.
Table 6.12 shows that in the Brown Family Corpora, the diachronic distribution
of the case of personal pronoun subjects of absolute clauses is relatively even, with
nominatives taking the advantage over accusatives. Except for one more nomina-
tive personal pronoun in the 1961 corpora, there is no other historical difference.

Stylistic Distribution

According to the statistics in Sects. 6.2.1.3 and “Stylistic Distribution of Personal


Pronoun Subject”, there are significant differences in stylistic distributions of
absolute clauses and of personal pronoun subjects. In this section, we will focus
on the stylistic distribution of the case of personal pronoun subjects.
Table  6.13 shows that in terms of the case-marked absolute clauses with the
first person and third person pronouns as subject, except learned in which there
are no data, nominatives in all the other three styles take an overall advantage over
accusatives. Like the overall frequency distribution of absolute clauses, absolute
clauses with personal pronoun subject are mainly used in fiction. Nevertheless,
the original data cannot fully explain that nominative personal pronouns are more
competitive than accusative personal pronouns as subject of absolute clauses.
This is because the occurrences of personal pronouns differ greatly in different
­sub-corpora of different styles. Therefore, to facilitate effective comparison, the
original frequencies will also be converted into standard frequencies of per million
personal pronouns, as is shown in Table 6.14.

Table 6.12  Historical 1961 1991/1992 2009 Total


distribution of the case
Nominative 10 9 9 28
of personal pronoun subjects
of absolute clauses Accusative 7 7 7 21
Total 17 16 16 49

Table 6.13  Stylistic Press General Learned Fiction Total


distribution of the case prose
of personal pronoun subjects
Nominative 4 9 0 15 28
of absolute clauses
Accusative 1 8 0 12 21
Total 5 17 0 27 49
6.2  Research Based on Brown Family Corpora 121

Table 6.14  Stylistic distribution of the case of personal pronouns in the Brown Family Corpora
Press General prose Learned Fiction Total
Nominative 14,238 33,070 5,094 62,325 114,727
Accusative 5,596 16,891 2,376 33,896 58,759
Total 19,834 49,961 7,470 96,221 173,486

500

400

300

200

100

0
Learned Press General prose Fiction Average

Nominative 0 281 272 241 244


Accusative 0 179 474 354 357
Average 0 252 340 281 282

Fig. 6.7  Stylistic distribution of the case of personal pronoun subjects of absolute clauses


(per million words)

As can be seen, among these four styles, there are altogether 173,486 case-
marked first and third person pronouns, including 114,727 nominatives and 58,759
accusatives, the ratio being 1:0.512. In addition, in these four styles, nomina-
tives take an overall advantage over accusatives. Of the 49 case-marked personal
pronouns, nominatives and accusatives are of the ratio 1:0.75. According to the
overall occurrences of the case of personal pronouns, we can figure out the stylis-
tic distribution of the case of personal pronoun subjects of absolute clauses (see
Fig. 6.7).
Table 6.13 shows that the 49 case-marked personal pronouns are mainly in fic-
tion and general prose, and fiction exceeds general prose in number. This is mainly
because there are more occurrences of personal pronouns in fiction and general
prose than in the other two registers. However, as is shown in Fig. 6.7, after con-
verted into the standard frequency of per million personal pronouns, the scene has
greatly changed. In other words, general prose conversely stands out from fiction,
and even to such an extent that the average frequency of fiction is slightly lower
than the total average frequency. Figure 6.7 also shows that in different styles,
the distributions of nominatives and accusatives differ greatly. In the four styles,
accusatives take the significant advantage against nominatives. In general prose,
there is the widest gap between nominatives and accusatives, while in fiction,
122 6  Absolute Clauses Distributed in Three Corpora

the proportion of nominatives to accusatives keeps steady with the total average
­frequency. It is only in press that accusatives fall behind nominatives, and even in
press, there are more nominatives than in fiction. It should be noted that personal
pronouns as well as absolute clauses with personal pronoun subject are seldom
used in learned. In the six learned sub-corpora, there are only two absolute clauses
with itself as subject.
The above statistics indicates that when personal pronouns have to be used as
the subject of absolute clauses, accusative personal pronouns are more likely to be
taken rather than nominative personal pronouns.

6.2.4 Built-in Problems

From the above quantitative research, we can draw the following conclusion. The
distribution of absolute clauses manifests no distinct regional or historical differ-
ences in the Brown Family Corpora, but manifests significant stylistic differences.
Among the three types of expansion, the number of absolute clauses of extension
is the largest, while that of enhancement, the smallest. On the whole, the total
number of the absolute clauses of enhancement goes on a steady decline, while
that of extension keeps increasing. In terms of the stylistic distribution, absolute
clauses mainly occur in fiction, but seldom in learned. This is obviously not in
agreement with the traditional view that absolute clauses are mostly used in formal
texts, because from the perspective of formality, learned is the most formal, then
press, general prose and fiction follow in succession. However, among these four
styles, the distribution of absolute clauses is right the other way, with fiction the
most, and then come general prose, press, and learned in turn.
In order to demonstrate the feasibility of the analysis on stylistic distribution, a
comparison will be drawn between the overall frequency of absolute clauses and
the frequency of absolute clauses with personal pronoun subject. To facilitate the
effective comparison, we convert the number of absolute clauses with personal
pronoun subject from 61 to 382 (see Fig. 6.8).
Figure 6.8 shows that absolute clauses share almost the same stylistic distribution
trend with absolute clauses with personal pronoun subject. From learned to press
to general prose, both are on steady increase, but the increasing speed of absolute
clauses with personal pronoun subject is a bit higher. However, from general prose
to fiction, both increase dramatically, and the overall frequency of absolute clauses
increases faster than that of absolute clauses with personal pronoun subject. This
indicates that absolute clauses tend to be applied in fiction rather than in learned. In
addition, there are no spoken data in the Brown Family Corpora, and among these
four registers, fiction is the closest to spoken. However, among the four registers,
fiction is the richest in absolute clauses. Therefore, we have the first question to
consider.
6.2  Research Based on Brown Family Corpora 123

180
160
140
120
100
80
60
40
20
0
Learned Press General prose Fiction
Overall Frequency 20.08 27.47 32.77 166.67
Personal Pronoun 13.09 29.69 55.73 132.26

Fig. 6.8  Comparison between stylistic distributions of absolute clauses and absolute clauses


with personal pronoun subject

Q1: Can this distribution indicate that absolute clauses tend to be used in
­informal texts?
The above quantitative analysis shows that neither in terms of the overall fre-
quency nor in terms of the frequency of the personal pronoun subject, the distri-
bution of absolute clauses shows distinct regional or historical differences. With
respect to historical distribution, the span of time of the Brown Family Corpora is
no longer than 50 years, within which the evolution tendency of language is not
always sensitive. The historical distribution research of absolute clauses might not
have caught up with its own internal trend of historical variation. Therefore, we
have the second question here.
Q2: Can absolute clauses manifest historical differences in a longer span of time?
In order to answer the questions, we will observe language facts from the BNC
and COCA in the following sections.

6.3 BNC-Based Observations

In this section, we will use the BNC to carry out a further quantitative research
on the stylistic distribution of absolute clauses. The research is mainly about the
application of absolute clauses in both informal and formal English.
124 6  Absolute Clauses Distributed in Three Corpora

6.3.1 Overall Frequency

On the basis of the five regular expressions written in Chap. 4, we have


retrieved 1935 absolute clauses of expansion, including 114 absolute clauses of
­enhancement, 1,500 of extension, and 321 of elaboration.
As for the absolute clauses of projection, we collected only the collocations of
a noun or pronoun, a projection verb and a personal pronoun followed by an infini-
tive, present participle, or past participle. And we selected several most representa-
tive verbs that can project both finite clauses and absolute clauses, such as believe,
consider, expect, feel, find, hear, know, see, want. We then wrote the following
search queries:
SQ6. [nn*]|[p*] [v*] [pp*] [v?g*]|[v?n*]
SQ7. [nn*]|[p*] [v*] [pp*] to [v?i*]
Using these search queries, we collected 6,347 absolute clauses of projection
from the BNC (see Table 6.15).
Table 6.15 shows that absolute clauses in the BNC mainly occur in fiction and
do not tend to be used in either academic or magazine. This is in agreement with
the analysis of the stylistic distribution in the Brown Family Corpora and with the
traditional view that absolute clauses are formal and infrequent. Even in the infor-
mal spoken texts, absolute clauses of expansion are infrequently used, counting
the least among the seven styles. Therefore, it comes whether absolute clauses are
seldom used in either formal or informal texts. In other words, what a stylistic dis-
tribution pattern do absolute clauses abide by?
The original frequencies were converted into the standard frequencies of per
hundred million words (see Fig. 6.9).
It can be seen from Fig. 6.9 that like in the Brown Family Corpora, absolute
clauses of expansion tend to occur in fiction and they are seldom used in either
spoken or academic. Absolute clauses of projection, however, are mostly used in
fiction and spoken and are least used in academic. Fiction is of great advantage in
both expansion and projection.

Table 6.15  Distribution of absolute clauses in the seven styles in BNC


Expansion Projection Total
Enhancement Extension Elaboration Sub-total
Spoken 3 27 6 36 1,135 1,171
Newspaper 8 36 1 45 411 456
Magazine 16 89 24 129 293 422
Non-academic 17 92 48 157 442 599
Miscellaneous 26 186 46 258 836 1,094
Academic 7 36 32 75 298 373
Fiction 37 1,034 164 1,235 2,932 4,167
Total 114 1,500 321 1,935 6,347 8,282
6.3  BNC-Based Observations 125

20000
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
Spoken Newspaper Magazine Non-ac Misc Academic Fiction
Expansion 361 430 1776 952 1238 489 7763
Projection 11391 3927 4035 2680 4012 1944 18429

Fig. 6.9  Stylistic distribution of absolute clauses in BNC (per hundred million words)

To illustrate the reliability of the stylistic distribution of absolute clauses, a


comparison will be conducted between the Brown Family Corpora and the BNC.
In the BNC, after converted into the standard frequency of per hundred million
words, the numbers of absolute clauses of expansion and projection are 13,009
and 46,418, respectively. However, as is discussed in Chap. 4, we take such styles
of magazine, non-academic and miscellaneous in the BNC as general prose in
the Brown Family Corpora, and the standard frequencies of expansion and pro-
jection in general prose are 1,220 and 3,523 per 100 million words, respectively.
Therefore, after being integrated into five styles such as spoken, newspaper, gen-
eral prose, academic, and fiction, the total numbers of expansion and projection
are 10,263 and 39,214, respectively. To facilitate comparison, we then convert the
total number of expansion into that of projection. Because there are no data of
spoken in the Brown Family Corpora, we take the total number of projection in
the four styles of newspaper, general prose, academic, and fiction, i.e., 27,823, as
the base number, and then convert the numbers of expansion and projection in the
Brown Family Corpora. Then, the four sets of data are compared (see Fig. 6.10).
Figure  6.10 shows that absolute clauses in these two corpora share almost the
same stylistic distribution trend. That is, the numbers of absolute clauses in both the
two corpora are on steady decrease from fiction to general prose, newspaper, and
academic except that of projection in the BNC where the number in newspaper
ranks the second among these four styles and that of expansion in the BNC where
the number in newspaper is slightly smaller than that in academic. When taking the
style of spoken into consideration, however, we find that the number of expansion
in spoken in the BNC is the smallest among the five styles, while that of projec-
tion is much larger, ranking the second among the five styles. It can be inferred that
compared with the other three styles, spoken and fiction prefer projected non-finite
126 6  Absolute Clauses Distributed in Three Corpora

35000

30000

25000

20000

15000

10000

5000

0
Spoken Newspaper General prose Academic Fiction
Brown Expansion 3089 3700 2360 18674
Brown Projection 4284 4309 861 18368
BNC Expansion 1381 1643 4661 1869 29661
BNC Projection 11391 3927 3523 1944 18429

Fig. 6.10  Comparison of stylistic distribution of absolute clauses between the Brown Family


Corpora and BNC

clauses to finite clauses. It is also noticed that in the Brown Family Corpora, there
are much fewer absolute clauses of projection in academic than those of expansion
and those of expansion and projection in the BNC. This is because we collected only
the absolute clauses with personal pronoun subject and pronouns are not preferred
in academic. This is also true for absolute clauses of projection in the BNC although
the gap is not that sharp. The four distribution lines in Fig. 6.10 are nearly over-
lapped, but some obvious differences can still be noticed. For instance, the number
of expansion in fiction in the BNC is larger than that in the Brown Family Corpora.
This comparison can at least explain that the frequency of absolute clauses in
the Brown Family Corpora and that of projection in the BNC are the lowest in for-
mal academic, and it is only slightly higher than that in spoken and that in news-
paper in the BNC. According to the data in the BNC, although absolute clauses do
not occur considerably in the formal academic, they are really rarely seen in the
informal spoken texts.

6.3.2 Functions

Although absolute clauses of elaboration are not included in the traditional sense
of absolute clauses, this kind of construction can also be considered as shrunken
finite clauses as other two function types of absolute clauses. To guarantee the
validity of the data analyses, we take absolute clauses of extension and enhance-
ment and those of projection in the BNC as examples to analyze the tendency to
choose absolute clauses from attendant circumstances introduced by with, from
6.3  BNC-Based Observations 127

finite adverbial clauses connected by conjunctions, and from complement clauses


introduced by the complementizer that. It is obvious that not all the constructions
with with can form absolute clauses. According to traditional grammar, absolute
clauses are separated by punctuation marks from the main clauses, and usually
attendant circumstances introduced by with are also separated by punctuation
marks. Therefore, we will retrieve all the collocations with with after punctuation
marks from the BNC, totaling 71,954 (see Table 6.16).
Likewise, not all constructions linked by conjunctions are finite adverbial
clauses, and some conjunctions can also be used as prepositions. Relevant limi-
tations thus should be set on the retrieval of conjunctions. We therefore retrieve
the collocations linked by the most commonly used subordinating conjunctions
(i.e., when, if, although, before, while, though, because) from the BNC. Such
restriction cannot guarantee the possibility of the role change from conjunction to
preposition, but it helps screen out most of the relative adverbs introducing relative
clauses (see Table 6.17).
As for the absolute clauses of projection, we collect complement clauses intro-
duced by the complementizer that. To extract the that clauses, we write the regular
expression 8 under the same condition as regular expressions 6 and 7 still using
the most representative verbs believe, consider, expect, feel, find, hear, know, see,
and want (Table 6.18).
SQ8. [nn*]|[p*] [v*] [pp*] that [v?g*]|[v?n*]
To facilitate comparison, the original frequency is converted into the standard
­frequency of per hundred million words (see Tables 6.19, 6.20 and 6.21).
Thereafter, with the total frequencies of with, subordinators and that in per
hundred million words, we observe the number of absolute clauses in each of
the five registers and analyze the tendency to form absolute clauses of attendant

Table 6.16  Stylistic distributions of with and absolute clauses of extension


Spoken Newspaper General prose Academic Fiction Total
With construction 2,362 7,823 39,133 10,481 12,155 71,954
Absolute clauses 27 36 367 36 1,034 1,500

Table 6.17  Stylistic distributions of subordinators and absolute clauses of enhancement


Spoken Newspaper General prose Academic Fiction Total
Subordinators 24,204 18,498 119,617 43,471 42,481 248,262
Absolute clauses 3 8 59 7 37 114

Table 6.18  Stylistic distributions of complementizer that and absolute clauses of projection


Spoken Newspaper General prose Academic Fiction Total
That clauses 4,226 5,043 5,212 2,635 5,078 22,194
Absolute clauses 1,135 411 1,571 298 2,932 6,347
128 6  Absolute Clauses Distributed in Three Corpora

Table 6.19  Stylistic distributions of with and absolute clauses of extension (per hundred million
words)
Spoken Newspaper General prose Academic Fiction Total
With construction 23,706 74,744 87,757 68,362 76,402 330,971
Absolute clauses 271 344 823 235 6,499 8,172

Table 6.20  Stylistic distributions of subordinators and absolute clauses of enhancement (per


hundred million words)
Spoken Newspaper General prose Academic Fiction Total
Subordinators 242,923 176,737 268,246 283,537 267,020 1,238,463
Absolute clauses 30 76 132 46 233 517

Table 6.21  Stylistic distributions of complementizer that and absolute clauses of projection (per


hundred million words)
Spoken Newspaper General prose Academic Fiction Total
That clauses 42,414 48,183 11,688 17,187 31,918 151,390
Absolute clauses 11,391 3,927 3,523 1,944 18,429 39,214

circumstances introduced by with (Figs. 6.11 and 6.12), finite adverbial clauses


introduced by subordinators and complement clauses introduced by that from
the BNC (Figs. 6.13, 6.14, and 6.15).
Figure 6.11 shows that under the condition that the total number of with con-
structions is equal to that of absolute clauses of extension, among those five styles,
only in fiction absolute clauses outnumber with constructions, while in the other
four styles, with constructions take advantages in number. The difference between
the two is the smallest in spoken texts (see Fig. 6.12).
Figure  6.12 shows that although the absolute number of absolute clauses of
extension is the smallest in spoken texts, the number of with constructions is also
the smallest among the five styles. By contrast, the ratio of absolute clauses of
extension to with constructions in spoken texts is higher than that in newspaper,
general prose, and academic texts, with academic the lowest.
From Fig. 6.13, we can see that, when the total number of subordinators is
equal to that of absolute clauses of enhancement, absolute clauses outnumber sub-
ordinators in newspaper, general prose, and fiction, respectively. As to subordina-
tors, they outnumber absolute clauses in spoken and academic texts (see Fig. 6.14
for the ratio of absolute clauses to subordinators in each of the five styles).
According to the data in Fig. 6.14, absolute clauses of enhancement are rather
preferable in newspaper, general prose, and fiction, while enhancing clauses intro-
duced by subordinators are preferable in spoken and academic. Besides, in spoken
texts, the absolute number of absolute clauses of enhancement is the smallest in per
hundred million words, and their ratio to subordinators is also the lowest. This sug-
gests that absolute clauses of enhancement are the least preferable in spoken texts.
6.3  BNC-Based Observations 129

300000

250000

200000

150000

100000

50000

0
Spoken Newspaper General prose Academic Fiction
with construction 23706 74744 87757 68362 76402
Absolute clauses 10976 13932 33332 9518 263213

Fig. 6.11  Tendency of with constructions to form absolute clauses of extension

400

350

300

250

200

150

100

50

0
Spoken Newspaper General prose Academic Fiction
Ratio 46.3 18.64 38 14 345

Fig. 6.12  The ratio of the absolute clauses of extension to with constructions (equal totality)

Figure 6.15 shows that when the total number of absolute clauses of projection
is converted to that of complement clauses introduced by the complementizer that,
absolute clauses outnumber that clauses in spoken, general prose, and fiction, with
the gap in fiction being the largest and that clauses outnumber absolute clauses in
newspaper and academic. Figure 6.16 shows the ratio of absolute clauses of pro-
jection to that clauses in each of the five styles.
In Fig. 6.16, we can see that absolute clauses of projection are rather prefer-
able in spoken, general prose, and fiction, while that clauses are preferable in
130 6  Absolute Clauses Distributed in Three Corpora

600000

500000

400000

300000

200000

100000

0
Spoken Newspaper General prose Academic Fiction
Subordinators 242923 176737 268246 283537 267020
Absolute clauses 71864 182056 316203 110192 558147

Fig. 6.13  Tendency of enhancing clauses to form absolute clauses of enhancement

250

200

150

100

50

0
Spoken Newspaper General prose Academic Fiction
Ratio 29.58 103.01 117.878 38.86 209.03

Fig. 6.14  Ratio of absolute clauses of enhancement to subordinators (equal totality)

newspaper and academic. This suggests that absolute clauses of neither expansion
nor projection tend to occur in formal academic.
In terms of the two function types of absolute clauses distinguished by tradi-
tional grammar, corpus statistics shows that the number of absolute clauses of
enhancement is significantly smaller than that of extension. Since there are no
explicit conjunctions, the logico-semantic relationship of absolute clauses with
the primary clauses is sometimes unclear, and hence, there may be different
6.3  BNC-Based Observations 131

80000

70000

60000

50000

40000

30000

20000

10000

0
Spoken Newspaper General prose Academic Fiction
that clauses 42414 48183 11688 17187 31918
Absolute clauses 43976 15161 13601 7505 71147

Fig. 6.15  Tendency of that clauses to form absolute clauses of projection

250

200

150

100

50

0
Spoken Newspaper General prose Academic Fiction
Ratio 103.68 31.47 116.37 43.67 222.9

Fig. 6.16  Ratio of absolute clauses of projection to that clauses (equal totality)

interpretations. Putting the two function types of absolute clauses together, we will
discuss their overall stylistic distributions.
We refer to both with and subordinators as linkers. Then, the ratio of absolute
clauses to linkers in each style will be figured out (see Table 6.22; Fig. 6.17).
We can see from Table 6.22 and Fig. 6.17 that, with equal frequencies, absolute
clauses outnumber linkers only in fiction. In the other four styles (spoken, newspaper,
general prose, and academic), absolute clauses are fewer than linkers. These results
show that absolute clauses differ significantly in stylistic distribution. According to
132 6  Absolute Clauses Distributed in Three Corpora

Table 6.22  Stylistic distributions of linkers and absolute clauses


Spoken Newspaper General prose Academic Fiction Total
Linkers 26,566 26,321 158,750 53,952 54,636 320,216
Absolute clauses 30 44 426 43 1,071 1,614

450
400
350
300
250
200
150
100
50
0
Spoken Newspaper General prose Academic Fiction
Ratio 22.4 33.17 53.24 15.81 388.91

Fig. 6.17  Ratio of absolute clauses to linkers (equal totality)

the absolute number and the ratio to linkers, absolute clauses tend to occur in fiction
the most, but they are not preferable in academic. Since there are fewer hypotactic
clauses of extension and enhancement in spoken, absolute clauses are small in num-
ber in spoken. Non-finite projected proposal is preferable in spoken than in newspa-
per and academic, so there are more absolute clauses of projection in spoken than in
newspaper and academic. This will be further discussed in Chap. 7.

6.4 COHA-Based Observations

The Brown Corpora-based quantitative research shows no significant difference of


historical distribution within the span of time of nearly 50 years. In this section, we
will conduct a diachronic research based on COHA which covers a much longer
span of time, including overall frequency, and function types and case choice, etc.

6.4.1 Functional Distribution of Overall Frequency

On the basis of the regular expressions we wrote, we retrieved 8,577 absolute


clauses of expansion and 41,009 absolute clauses of projection (see Table 6.23).
6.4  COHA-Based Observations 133

Table 6.23  Historical distribution of the absolute clauses in COHA


Expansion Projection Total
Enhancement Extension Elaboration Sub-total
1810s 2 2 8 12 36 48
1820s 45 18 15 78 388 466
1830s 105 63 51 219 728 947
1840s 81 98 41 220 911 1,131
1850s 138 122 93 353 1,128 1,481
1860s 96 152 79 327 1,264 1,591
1870s 112 148 49 309 1,605 1,914
1880s 70 154 68 292 1,780 2,072
1890s 108 182 34 324 1,847 2,171
1900s 80 197 51 328 2,397 2,725
1910s 63 174 46 283 2,707 2,990
1920s 72 326 86 484 2,897 3,381
1930s 49 351 68 468 2,626 3,094
1940s 39 336 106 481 2,549 3,030
1950s 24 362 102 488 2,747 3,235
1960s 28 437 154 619 2,849 3,468
1970s 26 557 95 678 2,750 3,428
1980s 21 481 138 640 2,907 3,547
1990s 30 695 195 920 3,538 4,458
2000s 24 853 177 1,054 3,355 4,409
Total 1,213 5,708 1,656 8,577 41,009 49,586

In terms of the overall frequency of expansion in COHA, there are much more abso-
lute clauses of extension than those of enhancement and elaboration. Such a ­functional
distribution matches up in all the three corpora, i.e., the Brown Family Corpora,
the BNC and COHA. The number of absolute clauses of enhancement in each of the
three corpora is the smallest of the three types of expansion (see Table 6.24).
To facilitate effective comparison, we convert the total numbers of absolute
clauses of expansion retrieved from the Brown Family Corpora and the BNC into
that (8,577) from COHA. We do not include projection because we use ­different
regular expressions to collect absolute clauses of expansion from the Brown
Family Corpora and from the BNC and COHA, while we use the same regular
expressions to collect absolute clauses of projection. As for absolute clauses of
projection, we convert the total numbers from the three corpora into the standard
frequencies of per hundred million words (see Figs. 6.18 and 6.19).
Figure 6.18 shows that under the condition of equal totality, the number of abso-
lute clauses of enhancement is the smallest in all the three corpora, and of the three
corpora, the number of enhancement is the smallest in the BNC, with the Brown
Family Corpora and COHA following. Absolute clauses of extension are entirely
different, with the BNC the most, then COHA, and the Brown Family Corpora the
134 6  Absolute Clauses Distributed in Three Corpora

Table 6.24  Functional distribution of absolute clauses


Expansion Projection Total
Enhancement Extension Elaboration Sub-total
Brown Corpora 53 237 92 382 533 915
BNC 114 1,500 321 1,935 6,347 8,282
COHA 1,213 5,708 1,656 8,577 41,009 49,586
Total 1,380 7,445 2,069 10,894 47,889 58,783

7000

6000

5000

4000

3000

2000

1000

0
Brown BNC COHA
Enhancement 1190 505 1213
Extension 5321 6649 5708
Elaboration 2066 1423 1656

Fig. 6.18  Functional distributions of absolute clauses of expansion in the three corpora

12000

10000

8000

6000

4000

2000

0
Brown BNC COHA
Projection 8670 6593 10113

Fig. 6.19  Functional distributions of absolute clauses of projection in the three corpora


6.4  COHA-Based Observations 135

least. Figure 6.19 shows that absolute clauses of projection appear mostly in COHA,
and the number is the smallest in the BNC. This distribution is similar to that of
enhancement, but contradicts that shown in Fig. 6.2, according to which absolute
clauses of projection show a trend of increase along time. The historical distribution
of absolute clauses of projection will be further discussed in Sects. 6.4.2 and 6.4.3.
Generally, the four function types of absolute clauses share similar distribution
trend. In Fig. 6.18, the three lines are nearly parallel, with enhancement in all the
three corpora the lowest and extension the highest. Therefore, absolute clauses
manifest no obvious functional differences in all the three corpora.
It is worth noting that the research on the historical distribution based on the
Brown Family Corpora shows that over a span of time of less than 50 years, the
number of absolute clauses of extension is on the increase, while that of enhance-
ment on the decrease. Comparatively, among these three corpora, the data in
the BNC cover the shortest distance from nowadays, while those in COHA the
greatest distance. Coincidentally, in COHA absolute clauses of enhancement and
projection are the most in number and those of extension the fewest among the
three corpora. However, in the BNC, the scene is right opposite. Absolute clauses
of enhancement and projection are the fewest and those of extension, the most.
In the following, a research on historical distribution of absolute clauses retrieved
from COHA will be conducted to test whether absolute clauses of enhancement and
projection are decreasing in number and those of extension increasing along time.

6.4.2 Historical: Overall Frequency

Figure  6.20 shows the historical distribution of the 8,577 absolute clauses of
expansion and the 41,009 absolute clauses of projection given in Table 6.23 over
20 phases totaling a span of 200 years in COHA. To facilitate comparison, we con-
vert the number of expansion into that of projection.
Figure 6.20 shows that the absolute number of absolute clauses of both expan-
sion and projection in COHA is increasing, with only slight declines on several
phases. However, such an overall distribution can hardly manifest the real evolu-
tion of absolute clauses, for the total vocabulary of COHA is around 406 million
and it keeps changing over phases. The frequency over the phases is converted to
the standard frequency of per hundred million words (see Fig. 6.21) so as to facili-
tate comparison.
Figure  6.21 shows that when the original frequencies are converted into the
standard frequency of per hundred million words, the total number of absolute
clauses has increased. The historical distribution of absolute clauses of projection
is nearly uniformly increasing, while that of expansion is generally increasing with
some fluctuations in several phases. The distribution in the phases from 1810s to
1910s is irregular: The first four phases change alternately, while the last seven
phases are on a steady decline. The phases from 1920s to 2000s are on a grad-
ual increase, during which the three phases from 1920s to 1950s keep flat and the
136 6  Absolute Clauses Distributed in Three Corpora

6000

5000

4000

3000

2000

1000

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Expansion 57 373 1047 1052 1688 1563 1477 1396 1549 1568 1353 2314 2238 2230 2333 2960 3242 3060 4399 5039

Projection 36 388 728 911 1128 1264 1605 1780 1847 2397 2707 2897 2626 2549 2747 2849 2750 2907 3538 3355

Fig. 6.20  Overall historical distribution of absolute clauses in COHA (equal totality)

18000
16000
14000
12000
10000
8000
6000
4000
2000
0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s
Expansion 4826 5385 7601 6555 10248 9165 7957 6871 7519 7096 5960 9020 9097 9159 9505 12345 13613 12087 15744 17042
Projection 3048 5601 5285 5677 6848 7411 8647 8762 8966 10847 11925 11293 10674 10469 11192 11882 11547 11489 12662 11347

Fig. 6.21  Overall historical distribution of absolute clauses in COHA (per hundred million


words)

phrase of 1980s decreases. Such a distribution can at least suggest that, absolute
clauses manifest a trend of increase rather than decrease in the evolution of nearly
two hundred years.

6.4.3 Function Types

Then how about the historical distributions of the functions of absolute clauses?
Since the total number of words in the phase of 1810s is the smallest among the 20
phases, the absolute number of absolute clauses is also the smallest accordingly,
and the function distribution of absolute clauses manifests no distinct regularity.
Absolute clauses do show a regular historical distribution from the phase of 1820s.
Therefore, the point of view in the traditional sense that absolute clauses are on a
gradual decrease is only applicable to those absolute clauses of enhancement. As
we can see, the number of absolute clauses of extension is obviously increasing
6.4  COHA-Based Observations 137

16000

14000

12000

10000

8000

6000

4000

2000

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Enhancement 810 3106 3645 2413 4006 2691 2885 1647 2507 1731 1327 1342 952 766 468 558 522 397 513 388

Extension 810 1242 2187 2920 3541 4261 3812 3624 4224 4263 3665 6076 6821 6598 7052 8714 11183 9084 11893 13794

Elaboration 3238 1035 1770 1222 2700 2215 1262 1600 789 1103 969 1603 1322 2082 1987 3071 1907 2606 3337 2862

Projection 3048 5601 5285 5677 6848 7411 8647 8762 8966 10847 11925 11293 10674 10469 11192 11882 11547 11489 12662 11347

Fig. 6.22  Historical distributions of functions of absolute clauses in COHA (per hundred


­million words)

(see Fig. 6.22). Even absolute clauses of enhancement themselves show no sign


of dying out. For a clearer visual presentation, we will also convert the number of
expansion into that of projection and the original frequencies into the standard fre-
quency of per hundred million words (see Fig. 6.22).
Figure 6.22 shows that from 1950s to 2000s, the number of absolute clauses of
enhancement maintains flat. This proves the result of the observations based on
the Brown Family Corpora and explains the increasing trend of extension and pro-
jection and the decreasing trend of enhancement shown in Fig. 6.18.

6.4.4 The Case of Personal Pronoun Subject

In Sect. 6.2.3, we studied the case choice of personal pronoun subjects of absolute


clauses based on the Brown Family Corpora. However, due to the small number
of words in the corpora, there are only a limited number of absolute clauses with
personal pronoun subject. What is more, with such a short span of time, the result
can hardly explain the diachronic changes of the case of personal pronoun sub-
jects. In this section, we will carry out a COHA-based historical research on the
absolute clauses with personal pronoun subject. There are only absolute clauses of
projection with accusative personal pronoun as subject, so we discuss only abso-
lute clauses of expansion (see Table 6.25).

6.4.4.1 Historical Distribution of Case

Among the 8,577 absolute clauses of expansion, there are 936 with pronoun
­subjects, of which 517 common pronouns, 419 personal pronouns. And of the
­personal pronouns, there are 289 with case markers, including 228 nominatives
138 6  Absolute Clauses Distributed in Three Corpora

Table 6.25  Case distribution of personal pronoun subjects of absolute clauses in COHA


Time Common Personal Nominative Accusative Absolutive Total
1810s 0 0 0 0 0 0
1820s 3 4 1 0 3 7
1830s 16 39 16 1 22 55
1840s 19 0 0 0 0 19
1850s 33 32 21 2 9 65
1860s 46 16 10 2 4 62
1870s 14 6 6 0 0 20
1880s 48 18 16 1 1 66
1890s 37 29 24 1 4 66
1900s 13 5 4 0 1 18
1910s 23 9 7 1 1 32
1920s 41 49 30 7 12 90
1930s 23 27 15 6 6 50
1940s 21 16 9 3 4 37
1950s 39 14 5 3 6 53
1960s 28 52 36 7 9 80
1970s 33 27 11 4 12 60
1980s 19 25 12 4 9 44
1990s 26 28 3 10 15 54
2000s 35 23 2 9 12 58
Total 517 419 228 61 130 936

1.2 Nominative
Accusative
1

0.8

0.6

0.4

0.2

0
1820s 1830s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 6.23  Proportions of the case distribution of absolute clauses with personal pronoun subject
in COHA

and 61 accusatives. The proportions of the case-marked personal pronoun subjects


in each phase are shown in Fig. 6.23.
Except that there are no data collected in the phases of 1810s and 1840s, over
the evolution of 200 years, in terms of the case of absolute clauses with personal
pronoun subject, nominatives are in a decreasing trend, and accusatives an increas-
ing trend. Before 1900s, nominatives dominate accusatives, and there are even no
6.4  COHA-Based Observations 139

accusatives collected in the phases of 1820s, 1860s, and 1900s. However, after
the phase of 1900s, nominatives start to decrease and gradually lose the dominant
position, while accusatives begin to increase and outnumber nominatives after
the phase of 1990s and take over the dominant position. Although such a result is
not manifested in the Brown Family Corpora-based research, the Brown Family
Corpora-based research does not manifest the dominance of nominative pronouns.
Instead, nominatives and accusatives are evenly distributed along time, with nomi-
native pronouns on a decreasing trend.
Researches based on both the Brown Family Corpora and COHA show that
absolute clauses with accusative pronoun subject are not wrong at all; and they
are on a significant increase in comparison with absolute clauses with nominative
pronoun subject.

6.4.4.2 Functional Distribution of Case

In this section, we will analyze the functional distributions of absolute clauses


with case-marked personal pronoun subject in COHA (see Table 6.26).
Table 6.26 shows that among the three function types of absolute clauses, the
accusative personal pronoun subjects are preferable in extension, accounting for
almost half of the nominative pronoun subjects, but seldom in enhancement and
elaboration. To facilitate effective comparison, we convert the frequencies of both

Table 6.26  Functional Enhancement Extension Elaboration Total


distribution of the case of
Nominative 34 99 95 228
absolute clauses
Accusative 5 46 10 61
Total 39 145 105 289

7000

6000

5000

4000

3000

2000

1000

0
Enhancement Extension Elaboration
Nominative 1279 3724 3574
Accusative 703 6468 1406
Overall 1213 5708 1656

Fig. 6.24  Functional distribution of the case of absolute clauses


140 6  Absolute Clauses Distributed in Three Corpora

nominatives and accusatives into the standard frequency of 8,577 and analyze the
functional distributions of personal pronoun subjects, as is shown in Fig. 6.24.
Figure 6.24 shows that in comparison with the overall frequency, nominatives
are preferable in absolute clauses of elaboration and enhancement, while accusa-
tives are preferable in absolute clauses of extension.

6.4.4.3 Functional Distribution of Case Over Time

The above research on the overall historical distribution of absolute clauses with
case-marked personal pronoun subject shows that the accusatives are on a general
trend of increase, and nominatives, decrease. In this section, we will discuss the
functional distributions of the case-marked personal pronoun subjects of absolute
clauses over time (see Table 6.27; Figs. 6.25, 6.26, and 6.27).
Table 6.27 and Figs. 6.25, 6.26, and 6.27 show that the historical distribution of
the case of absolute clauses has no significant differences in the three functions.
Although accusative pronouns count few among all the three functions, they tend
to replace nominative pronouns.

Table 6.27  Functional distribution of the case of absolute clauses over time


Enhancement Extension Elaboration Total
Nom. Accus. Sub-total Nom. Accus. Sub-total Nom. Accus. Sub-total
1810s 0 0 0 0 0 0 0 0 0 0
1820s 0 0 0 1 0 1 0 0 0 1
1830s 5 1 6 3 0 3 8 0 8 17
1840s 0 0 0 0 0 0 0 0 0 0
1850s 6 0 6 8 2 10 7 0 7 23
1860s 3 0 3 4 2 6 3 0 3 12
1870s 1 0 1 4 0 4 1 0 1 6
1880s 3 0 3 11 1 12 2 0 2 17
1890s 3 1 4 12 0 12 9 0 9 25
1900s 1 0 1 2 0 2 1 0 1 4
1910s 2 0 2 1 1 2 4 0 4 8
1920s 2 0 2 20 7 27 8 0 8 37
1930s 1 0 1 7 5 12 7 1 8 21
1940s 1 0 1 2 3 5 6 0 6 12
1950s 1 0 1 4 2 6 0 1 1 8
1960s 3 2 5 10 5 15 23 0 23 43
1970s 1 0 1 5 4 9 5 0 5 15
1980s 1 0 1 3 3 6 8 1 9 16
1990s 0 0 0 1 6 7 2 4 6 13
2000s 0 1 1 1 5 6 1 3 4 11
Total 34 5 39 99 46 145 95 10 105 289
6.4  COHA-Based Observations 141

Nominative Accusative

0.8

0.6

0.4

0.2

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 6.25  Historical distribution of the case of absolute clauses of enhancement

Nominative Accusative
1

0.8

0.6

0.4

0.2

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 6.26  Historical distribution of the case of absolute clauses of extension

Nominative Accusative
1

0.8

0.6

0.4

0.2

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 6.27  Historical distribution of the case of absolute clauses of elaboration

The COHA-based research shows that the total number of absolute clauses is
not decreasing but increasing along time. The opinion held by traditional gram-
marians that absolute clauses are on a gradual decrease is true only for those abso-
lute clauses of enhancement, but the number of absolute clauses of extension is
obviously increasing. Even those of enhancement are not disappearing, for the dis-
tribution of absolute clauses of enhancement tends to stay at the same level in the
last 50 years.
142 6  Absolute Clauses Distributed in Three Corpora

6.5 Summary

The Brown Family Corpora-based research does not show significant difference
of regional distribution of absolute clauses. The COHA-based research shows that
over the span of 200 years, the total number of absolute clauses tends to increase
but not decrease. This is mainly manifested in the obvious increase of absolute
clauses of extension. Meanwhile, although the number of absolute clauses of
enhancement is decreasing along time, it is by no means disappearing. The data
of the last 50 years show that the distribution of absolute clauses of enhancement
tends to be leveling off. The reason may be the tendency of grammaticalization of
absolute clauses: being fixed into stereotyped expressions.
Both the Brown Family Corpora-based research and the BNC-based research
show that the stylistic distributions of absolute clauses manifest significant differ-
ences. Among the four styles, i.e., press, general prose, learned, and fiction, in the
Brown Family Corpora, absolute clauses tend most to occur in fiction, and fiction
is the most informal and learned the most formal. However, the fact is that the fre-
quency of absolute clauses is the lowest in learned. This is not in agreement with
traditional grammar. Although the fact that absolute clauses are seldom used in the
typical informal style of spoken complies with traditional grammar, there are few
adverbial clauses introduced by conjunctions in spoken. That is to say, the reason
why absolute clauses are seldom used in spoken is that there are the fewest subor-
dinate clauses in spoken than in other styles.
Although the subject of absolute clauses may be a noun or a pronoun, absolute
clauses with pronoun subject are few in number, and the pronoun subjects show
significant differences in stylistic distribution. The Brown Family Corpora-based
research shows that personal pronoun subjects of absolute clauses are mainly
concentrated in general prose and fiction. However, personal pronoun subjects
of absolute clauses per million personal pronouns gain the highest frequency in
general prose, then in fiction and press in succession. In addition, from the per-
spective of case distribution, accusatives tend to occur in general prose and fiction,
and nominatives in press. The COHA-based research on the historical distribution
of case provides further evidence that the number of accusative personal pronoun
subjects is increasing along time.

References

Kortmann, B. (1991). Free adjuncts and absolutes in English: Problems of control and interpre-
tation. London: Routledge.
Stump, G. T. (1985). The semantic variability of absolute constructions. Dordrecht: D. Reidel
Publishing Company.
Zandvoort, R. W. (1972). A handbook of English grammar (6th ed.). London: Longman.
Chapter 7
Discussions

Form, function, case, stylistic characteristics, and historical evolution, etc., have
been studied by many linguists as far as absolute clauses are concerned. However,
there still exist some significant problems. In Chap. 6, we have conducted a
­corpus-based quantitative research on the three function types of absolute clauses,
focusing on the function, case, style, and time aspects. In this chapter, a discussion
will be conducted on the research results.

7.1 Function

In Sect. 6.4.2, the function types of absolute clauses have been discussed.


According to traditional grammar, the logical roles of absolute clauses can be clas-
sified into two groups: attendant circumstances and clausal adjuncts. However, in
actual language use, a third group can be distinguished, i.e., appositive. These three
function types can be explained by applying the clause complex theory in SFL. That
is, as non-finite clauses, absolute clauses, together with the matrix clauses, can form
clause complexes realizing expanded elaboration (appositive), extension (attendant
circumstance), and enhancement (clausal adjuncts). However, in addition to these,
absolute clauses can also be used alone as independent clauses, for example,
7-1a. I ran through my mental list of things I needed to check, knowing
I’d done everything already. Windows locked, post put on hold, boiler
switched off (CLOB_L).
b. The same look comes over their faces. Tears low (CORWN_M).
c. Urine only faintly positive. Her doctor reported that she was “a hermit type”
(LOB_J).
In 7-1, although the logical subject–predicate constructions are used on their
own, they are still semantically dependent on the precedent or following sentences

© Springer-Verlag Berlin Heidelberg 2015 143


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_7
144 7 Discussions

and form some logico-semantic relations with these sentences. In another sense,
these constructions have no finite verbs, nor are elliptical clauses formed out of the
consistency of syntactic structures, hence are non-finite clauses. In the mean time,
these constructions have their own subjects, hence are absolute clauses. In 7-1a,
the three absolute clauses function as appositives of the precedent clause I’d done
everything already, realizing elaboration. In 7-1b, the absolute clause is accompa-
nying the event of the precedent clause, realizing extension. In 7-1c, the absolute
clause is the reason of the following clause, realizing enhancement. The logico-
semantic relations realized by these absolute clauses are cohesive rather than
structural. However, these kind of absolute clauses are not included in the absolute
clauses in the traditional sense.
Using the regular expressions we wrote, we extracted 78, 281, and 997 absolute
clauses of expansion used alone from the Brown Family Corpora, the BNC, and
COHA, respectively (see Table 7.1).
Table 7.1 shows that in the three corpora, absolute clauses used alone account
for 11.07 % of the total. Although such absolute clauses do not form clause com-
plexes, they realize such logico-semantic relations of extension and enhancement
of time, reason, condition, concession, and so on, and they can also realize elabo-
ration. An independently used absolute clause construes a sequence together with
a relevant finite clause. The question is whether the number of absolute clauses
used independently is increasing or decreasing and whether the function distri-
bution of the independently used absolute clauses is consistent with that of the
dependently used absolute clauses (see Table 7.2).
To analyze the historical distribution of independently used absolute clauses,
we will convert the frequency in each phase into the standard frequency of per
hundred million words (see Fig. 7.1).
Figure  7.1 shows that the historical distribution of the independently used
absolute clauses is irregular and shows no clear tendency of increase or decrease.
The number of absolute clauses used dependently began to increase in 1910s and
increased rapidly after 1980s. Comparatively, the distance between the two distri-
bution lines became broader gradually after 1910s and broadened out rapidly after
1970s. Therefore, under the condition that the total number of absolute clauses is
on the increase, absolute clauses used independently are basically unchanged in
number, indicating that the number of the latter is relatively on the decrease.
According to the clause complex theory in SFL, independently used clauses
tend to be paratactic. Absolute clauses used independently are paratactic no matter

Table 7.1  Absolute clauses used independently


Clause complex Independent Total Percent of
Enhancement Extension Elaboration independent
Brown 53 237 92 78 460 16.96
BNC 114 1,500 321 281 2,216 12.68
COHA 1,213 5,708 1,656 997 9,574 10.41
Total 1,380 7,445 2,069 1,356 12,250 11.07
7.1  Function 145

Table 7.2  Absolute clauses used independently in COHA


Independent Dependent Total Percent of independent
1810s 2 12 14 14.29
1820s 4 78 82 4.878
1830s 12 219 231 5.195
1840s 16 220 236 6.78
1850s 43 353 396 10.86
1860s 65 327 392 16.58
1870s 39 309 348 11.21
1880s 23 292 315 7.302
1890s 14 324 338 4.142
1900s 38 328 366 10.38
1910s 18 283 301 5.98
1920s 73 484 557 13.11
1930s 35 468 503 6.958
1940s 48 481 529 9.074
1950s 52 488 540 9.63
1960s 62 619 681 9.104
1970s 137 678 815 16.81
1980s 74 640 714 10.36
1990s 130 920 1,050 12.38
2000s 112 1,054 1,166 9.605
Total 997 8,577 9,574 10.41

4000

3500

3000

2500

2000

1500

1000

500

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s
Independent 169 58 87 100 261 381 210 113 68 172 79 285 142 197 212 259 575 292 465 379
Dependent 1016 1126 1590 1371 2143 1917 1665 1437 1573 1484 1247 1887 1902 1976 1988 2582 2847 2528 3293 3565

Fig. 7.1  Proportion of absolute clauses used independently in COHA

what logico-semantic relations they realize. Clause complexes of enhancement


can be analyzed in double transitivity. An absolute clause of enhancement has the
potential to realize a figure in a clause complex and a circumstance in a simple
clause as well. Since absolute clauses have the potential to realize circumstances,
they have also the potential to realize participants. Therefore, in addition to the
146 7 Discussions

three function types, we also discussed absolute clauses functioning as subjects,


complements, and appositives. Whatever function it might be, the basic criterion
for identifying absolute clauses is that the subject of an absolute clause is not a
participant of the main verb in the primary clause. That is, the subject of an abso-
lute adjunct clause is not co-referential with that of the primary clause, the subject
of an absolute subject clause is not the subject of the main verb of the primary
clause, and the subject of an absolute complement clause is not the complement of
the main verb of the primary clause.
Like that of absolute adjunct clauses, the logico-semantic relations realized by
absolute subject clauses, absolute complement clauses, and absolute appositive
clauses are implicit, that is, they are not linked by any explicit conjunctive expres-
sions. The difference lies in that absolute adjunct clauses can realize such relations
as hypotactic extension and enhancement, and absolute subject clauses, absolute
complement clauses, and absolute appositive clauses realize only one certain rela-
tion each. The former should be recognized in context, and the latter, relatively
clear. This can be used to explain why conjunctions are essential in finite clauses
of hypotactic expansion to realize relator, but are not indispensible in finite clauses
of hypotactic projection and deep embedded finite clauses functioning as subjects
or appositives to realize relator. In order to further discuss the stylistic distribution
and the historical evolution of absolute clauses of subject and complement, we can
retrieve the corresponding non-finite clauses with subject from the corpora and
work out the proportion of such constructions and the corresponding finite clauses.
We use different regular expressions to retrieve different function types of abso-
lute clauses. In terms of absolute clauses of subject, the constructions composed of
nouns or pronouns and non-finite elements will be retrieved. The relation between
the nouns or pronouns in such constructions and the following main verbs can be
used to identify whether such constructions are absolute clauses or nominal groups
with post-modifier. The criterion of identification is that the nouns or pronouns in
such constructions are not the subjects of the main verbs, for example,
7-2a. People passing the farmhouse saw the smoke rising from the chimney
(BNC_FIC).
b. Children having fun is quite a structured exercise (BNC_MAG).
In 7-2a, People in the noun + participle construction People passing the farm-
house is the subject of the main verb saw, so the participle phrase passing the
farmhouse is the post-modifier of People. In 7-2b, Children in the noun + partici-
ple construction Children having fun is not the subject of the main verb is, but the
whole construction itself is the subject of is, so this construction forms an absolute
clause of subject.
In terms of absolute clauses of complement, the constructions composed of
nouns or pronouns and non-finite elements following relevant projection verbal
verbs, projection mental verbs, and some relational verbs will be retrieved. The
criterion of identification is that the nouns or pronouns in such constructions are
not the complements of the main verbs. In fact, absolute clauses of complement
are easier to be identified than those of subject. This is because the identification
7.1  Function 147

of absolute clauses of complement depends on the main verbs of the primary


clauses and that of subject on the meaning, having nothing to do with the main
verbs of the primary clauses. For example,
7-3a. Avis heard the water running in the sink (COCA_FIC).
b. She invited me to join the Science Club (COCA_FIC).
Since what we hear is always the sound of something rather than the something
itself, in 7-3a, the whole non-finite clause the water running in the sink instead
of the subject the water of the non-finite clause functions as the complement of
heard. Likewise, what we invite is the person himself rather than the action he is
performing, so in 7-3b, the subject me of the non-finite clause is in the first place
the complement of the main verb of the primary clause invited. This can also be
used to explain why the non-finite verbal group in 7-3a cannot be infinitive, and
why the non-finite clause in 7-3b cannot be present participle. Compare
7-4a. I began to study photography and film in college at Cooper Union
(COCA_MAG).
b. I began studying French in high school (COCA_FIC).
According to traditional grammar, both the two example sentences in 7-3
are constructions of subject + predicate + object + complement, meaning that
“although the predicate verb has its own object, the meaning of the sentence is not
complete, hence a complement is needed” (Huang 1998). However, the traditional
analysis is applicable to 7-3b but not to 7-3a. According to SFL, both the example
sentences in 7-3 are constructions of subject + main verb + complement, because
“the complement realizes a ‘situation’, which of course should be performed by
two or more elements” (Huang 1998). However, the functional analysis is appli-
cable to 7-3a but not to 7-3b. It should be noted that it is the main verb rather than
the form of the non-finite clause that determines to which element the subject of
the non-finite clause belongs. For example, the non-finite clauses in 7-5 are both
absolute clauses of complement.
7-5a. I believe you to be an honest fellow (COHA_FIC).
b. I imagine her being a BASE jumper when she’s 18 (COHA_MAG).
In constructions composed of nouns or pronouns and non-finite verbal groups,
it rarely happens that the noun or pronoun is not the subject of the main verb of
the primary clause; hence, it is impossible to form absolute clauses. Most of them
are constructions of nouns or pronouns + post-modifiers. Or sometimes, the non-
finite verbs function as adjuncts and nouns or pronouns are the subjects of the main
verbs. Such is commonly seen before the phase of 1850s in COHA, for example,
7-6a. Margaret seeing Hash, was inconsiderate enough to speak to him
(COHA_FIC 1845).
b. Man possessing reason, has an innate desire for knowledge (COHA_NF
1841).
c. Pauline Having perceived Gustavus, makes signs to him (COHA_FIC 1830).
148 7 Discussions

Table 7.3  Case-marked personal pronoun subjects


Absolute clauses Case-marked personal pronoun Proportion of pronouns
Nominative Accusative Total
Brown 382 8 5 13 3.4
BNC 1,935 33 29 62 3.2
COHA 8,577 228 61 289 3.37
Total 10,894 269 95 364 3.34

7.2 Case

Statistics shows that there are only a small number of absolute clauses with
personal pronoun subject. Table 7.3 displays the quantity of the case-marked
­
­absolute clauses with personal pronoun subject retrieved from the three corpora
based on relevant regular expressions.
In the three corpora, the number of case-marked personal pronoun subjects
accounts for 3.34 % of the total of absolute clauses and is uniformly distributed.
The propositions in the three corpora are very close. However, the propositions
of accusative pronouns to nominative pronouns in the three corpora are quite
­different (see Fig. 7.2).
According to the previous analysis, this is mainly because before the middle
of the twentieth century, absolute clauses gave priority to nominative personal
pronoun subjects in COHA, with quite few accusative pronouns. After 1960s,
­accusative pronouns began to increase rapidly in number.
The reason why there are so few absolute clauses with personal pronoun
­subject is that the subjects of absolute clauses are not co-referential with those of
the primary clauses and are less dependent on the subjects of the primary clauses.
The purpose of using absolute clauses is to avoid the subjects of the two clauses
referring to the same person or thing (Onions 1905 [2010]). Personal pronouns
do not carry semantic content and they usually refer back to an antecedent in

50
45
40
35
30
25
20
15
10
5
0
Brown BNC COHA
Accusative 38.46 46.77 21.11

Fig. 7.2  Proportion of accusatives to case-marked personal pronoun subjects


7.2 Case 149

the precedent context. For example, of the 28 absolute clauses with nominative
­personal pronoun subject retrieved from the Brown Family Corpora, there are five
he and five she, for example
7-7a. There they continued their studies at the university, she in art, he in archi-
tecture (BROWN_A).
b. At the end, Ruth and the Indian have exchanged cultural identities, she with
an Indian name and he, a Jewish one (FROWN _G).
The subjects of such absolute clauses are co-referential with those of the
p­ rimary clauses and they form a contrast. Neither of the personal pronoun sub-
jects is completely co-referential with that of the primary clause. In 7-7a, the per-
sonal pronouns also imply that the subject they of the primary clause contains a
male and a female, and in 7-7b, the he provides the information that the Indian in
the primary clause is a male. Certainly, these two personal pronouns can also be
reduced to the corresponding nouns, failing to avoid simple lexical repetition.
7-8a. The wife, Amra, and her lover are both savagely portrayed, she as incarnate
sensuality, “voluptuous” and “indolent”, possibly “a mischief maker”…
(BROWN_G).
b. And so they went, he choosing of all places an inn near Medmenham Abbey
(BROWN_K).
Although she in 7-8a and he in 7-8b do not form a contrast with other personal
pronouns, both are not co-referential with the subject of the primary clause, and
both convey some new information. For example, in 7-8a, she forms a contrast
with a part of the subject, her lover, so it can be recovered to the wife, Amra, and
in 7-8b, he implies that there is a male in they.
7-9a. The big brains—they more than doubled in size from Lucy’s—did not
appear until about 2 million years ago (FROWN_F).
b. We as Black people, we can control what we do to each other (Clob_G).
they in 7-9a and we in 7-9b are co-referential with the subjects of the primary
clauses. However, such absolute clauses are not grammatically acceptable. If the
subjects are removed, the former will become a normal non-finite clause and the
latter a prepositional phrase.
7-10 We’ll leave the car and work down five or six houses on foot, me on one
side, you on the other (CROWN_L).
Like 7-7, the subjects of the two absolute clauses in 7-10 refer back to the subject
of the primary clause, but the accusative subject me appears. Although you is not
case-marked, it can also be inferred as an accusative. Even if not, nominative and
accusative personal pronouns can also be used together sometimes, for example,
7-11 Both of us had gone out to Africa, shortly after Oxbridge, at the age of
25—he to edit Drum magazine, me to edit the Nyasaland Times (Clob_G).
150 7 Discussions

It can be seen that the case choice of personal pronoun subjects of absolute clauses
is arbitrary when there is a contrast. If there is not an antecedent in the precedent
context nor a contrast is formed, accusative personal pronouns will be first consid-
ered, for example,
7-12a. Me being a less classy guy than the Pope, I wolf whistled (Clob_N).
b. …every course of action…will result in the economy collapsing like a house
of cards in a hurricane, us all losing our jobs and homes… (Clob_G).
The main function of reflexive pronouns is to emphasize. Therefore, it is not
arbitrary to choose reflexive pronouns which cannot be replaced by nominative pro-
nouns. In the following example, if himself is changed into he, it will be co-referen-
tial with the subject of the primary clause, which is not grammatically acceptable.
7-13 As editor of the Criterion, Eliot at this period often used anthropological
material, himself selecting books for review and reading every word of what
would appear in print (BNC_ACAD).
Since reflexive pronouns always have antecedents, reflexive pronoun subjects
can be omitted when the function of emphasis is not foregrounded, and hence
form non-finite clauses, verbless clauses, or nominal groups.
7-14a. To this day, Hardy, himself having achieved the double of classical and
popular success as an actor, speaks of him at that time with unaffected
adulation (BNC_MISC).
b. After they had deposited their bags at the hotel, itself ramshackle and run-
down, they had gone on to the hospital (BNC_FIC).
c. The mother-of-pearl shimmers in the background of a Henry McFee, itself
a formula derived from Renoir (BROWN_J).
Sometimes, however, the removal of reflexive pronouns may lead to the change
or ambiguity of the logico-semantic relations, for example,
7-15a. “I suppose he does,” said Joan vaguely, herself nurturing a secret fondness
for that prince (BNC_FIC).
b. I made a rule that all gifts should be submitted for approval to the archi-
tect, himself a fine artist (LOB_G).
c. According to William Julius Wilson, of the University of Chicago, himself
black, poor blacks are even more isolated because any rich blacks who can
get out do so (BNC_MAG).
d. Now, his pupil and successor, Marcel Dupre, himself in his seventies and
a pioneer of organ records, has re-recorded it there in a coupling with
Widor’s fifth and “Gothic” symphonies (LOB_C).
Both in 7-15a and in 7-15b, the absolute clauses are adverbial clauses of cause.
If the reflexive pronouns are removed, they will be changed into a non-finite
clause of extension and a nominal group of appositive. Both in 7-15c and in 7-15d,
the absolute clauses are adverbial clauses of concession. If the reflexive pronouns
are removed, they will be changed into verbless clauses of elaboration.
7.2 Case 151

The personal pronoun subject of absolute clauses can be either nominative


or accusative. Neither the view that the reason why absolute case changed from
dative to nominative is the loss of case inflections nor the view that the use of a
noun in the zero-form or a pronoun in the subject form is a continuation of the
Old English usage with the noun before the participle in the zero case can give
a reasonable explanation for the presence of accusative personal pronoun subject
of absolute clauses. Ablative absolutes in Latin can usually be translated either
into constructions of with + noun + participle or into relevant adverbial clauses in
English, for example,
7-16 Eo- imperium tenente, ēventum timeo- (Wheelock and LaFleur 2005: 157).

a. With him holding the power, 
b. Since he holds the power,



c. When he holds the power, I fear the outcome.
d. If he holds the power,




e. Although he holds the power,

In traditional grammar, 7-16a is an augmented absolute. Grammatically, him


is the object of the preposition with, and the non-finite verbal group functions as
the logical predicate of him. 7-16b–e are adverbial clauses of cause, time, con-
dition, and concession, respectively. According to SFL, all the sentences in 7-16
are clause complexes, among which the secondary clause of 7-16a is a non-finite
clause, with the preposition with realizing relator. If the logico-semantic relation
of the non-finite clause is clear, an explicit conjunctive expression is not always
necessary to realize relator. Therefore, with can be omitted in 7-16a, forming an
absolute clause, and the personal pronoun subject of the absolute clause remains
accusative. The secondary clauses in 7-16b-e are all finite clauses, with conjunc-
tions realizing relator. As the subjects of the secondary clauses are not co-refer-
ential with those of the primary clauses, the secondary clauses can all be changed
into absolute clauses, and the personal pronoun subjects remain nominative.
Table 6.26 shows that accusative personal pronoun subjects are mostly used in
absolute clauses of extension, and seldom in absolute clauses of enhancement. Of
the five accusative subjects, four are reflexive pronouns, and only one accusative
personal pronoun in the real sense. Since absolute clauses of elaboration are not
the outcome of the ellipsis of with, it is not advisable to choose accusative per-
sonal pronoun subject.

7.3 Style

Observations based on the Brown Family Corpora and the BNC show that the sty-
listic distributions of absolute clauses are significantly different. Absolute clauses
tend to occur in the style of fiction, but are seldom used in learned or spoken styles.
152 7 Discussions

This is not in agreement with the traditional view that absolute clauses are formal
and infrequent.
Why are absolute clauses popularly used in fiction? This may be due to the fact
that fiction is characterized with dynamic narration and static description. This
characteristic can also be reflected in the choice of language form. The first sen-
tence of the English novel Pride and Prejudice is famous not only in its semantic
content but also in its particular language form which will in turn glorify its basic
semantic content. Absolute clauses as a substitute form of finite clauses can func-
tion as a means to foreground the semantic content.
According to SFL, absolute clauses have their transitivity structure, thematic
structure, and information structure, but no mood structure. The interpersonal
metafunction of language is the function to enact social processes through mean-
ing. In narration or description, a social process is enacted, and then, narration or
description continues. Absolute clauses are used to avoid the frequent enacts of
social processes.
When used in the paratactic relations, absolute clauses of extension add new
content to the primary clauses. The corresponding finite forms of such absolute
clauses are the continuing clauses in paratactic clause complexes. Absolute clauses
of extension and enhancement are all dependent clauses and hence are flexible
in position. This facilitates to highlight the propositions enacted by the primary
clauses. Without the use of absolute clauses, the primary information will be
­submerged into various secondary information, for example,
7-17 Billee wagged his tail appeasingly, turned to run when he saw that appease-
ment was of no avail, and cried (still appeasingly) when Spitz’s sharp teeth
scored his flank. But no matter how Spitz circled, Joe whirled around on his
heels to face him, mane bristling, ears laid back, lips writhing and snarling,
jaws clipping together as fast as he could snap, and eyes diabolically gleam-
ing—the incarnation of belligerent fear. So terrible was his appearance that
Spitz was forced to forego disciplining him; but to cover his own discomfi-
ture he turned upon the inoffensive and wailing Billee and drove him to the
confines of the camp (COHA_FIC).
As can be expected, the typical characteristics of learned texts are the wide use
of passive voices and nominalizations, which are effective methods to objectivize
the proposition. Nominalization is also a method to compact the information con-
strued by a clause into a nominal group. Because of the compact of information,
clause complexes are not encouraged to use in learned, nor are absolute clauses.
The small number of absolute clauses in learned texts realize hypotactic elabora-
tion, for the purpose of the learned texts is not to narrate or describe but to explain
or argue, for example,
7-18a. Five papers are now in circulation, two from consume groups, one from
managers, and two from medical organisations (BNC_ACAD).
b. The test subjects are shown three video clips involving the same simple
action: a hand grasping a teacup (CROWN_J).
7.3 Style 153

In 7-18a, the three absolute clauses are the explanations to Five papers.
In 7-18b, the absolute clause is used to explain the same simple action.
According to the data, absolute clauses of elaboration do not take the advan-
tage in learned over the other styles. This is because the subjects of most absolute
clauses of elaboration are personal pronouns, and personal pronoun subjects are to
be avoided as much as possible in learned texts. For example, of the 61 absolute
clauses with personal pronoun subject retrieved from the Brown Family Corpora,
only two inanimate third person singular reflexive pronouns itself occur in learned.
Therefore, absolute clauses of elaboration, even though small in number, are still
mostly appear in other styles than in learned, for example,
7-19a. Here’s something about Miss Leefolt: she not just frowning all the time,
she skinny (CROWN_K).
b. It looks real nice, them sitting there, just the two of them (BNC_FIC).
c. It was still there, they both thought, she with greed and apprehension, he
with alarm (BNC_FIC).
In learned, the subjects of absolute clauses of elaboration are mostly common
nouns or pronouns. Of the 32 subjects of absolute clauses of elaboration, there are
only three reflexive personal pronouns (itself) retrieved from academic in the BNC,
for example,
7-20 
On 26 December the upper house of the USSR Supreme Soviet, itself
inquorate, voted a formal end to the original treaty of union (BNC_ACAD).

7-21a. Three were misdiagnosed as having colitis; one in childhood, and two oth-
ers at age 38 and 42 years respectively (BNC_ACAD).
b. Play another sound effect—perhaps a door opening, keys jangling, a car
moving off, a bomb exploding, or a dog barking—and it becomes difficult
not to link the two sounds together and make them part of the same story
(BNC_ACAD).

7.4 Time

The COHA-based research shows that over a span of 200 years, the historical dis-
tribution of absolute clauses manifests a significant difference, and different func-
tions of absolute clauses have different evolution trends. Extension is increasing;
enhancement is gradually decreasing but keeps steady in recent dozens of years
(not dying out at all); elaboration shows no obvious historical changes.
Although traditional grammar (e.g., Quirk et al. 1985) insists that absolute
clauses are not frequently used in today’s English, some fixed expressions such
as Weather permitting, All things considered, There being… and It being… are
still popularly seen. Most of such stereotyped absolute clauses realize hypotactic
enhancement. Table 7.4 shows some of the most frequently occurring absolute
154 7 Discussions

Table 7.4  Absolute clauses Construction Number Proportion (%)


of enhancement in three
Brown There being… 13 24.53
corpora
All things considered 5 9.43
Total 53 100
BNC Weather permitting 28 24.56
Total 114 100
COHA Weather permitting 56 4.62
It being… 71 5.9
The weather being… 54 4.45
Total 1213 100

clauses of enhancement retrieved from the three corpora. Here, the frequencies
of absolute clauses will not be compared, because different regular expressions
have been used in different corpora. For example, the regular expressions can be
used to retrieve absolute clauses beginning with nouns or pronouns in the BNC,
and a­bsolute clauses composed of nouns or pronouns and present participles.
Therefore, such constructions as all things considered, there being… can be
retrieved from neither of the two corpora.
We will now take the fixed expressions (the) weather permitting/being…
retrieved from COHA as examples to analyze their historical distributions. The
data are shown in Table 7.5.
When combining the two expressions with the together and combining the
other two without the together, we have 57 instances with the and without the
each. To facilitate comparison, we will convert the frequencies with the and those
without the in the 20 phases into the standard frequencies of per hundred million
words (see Fig. 7.3).
Figure 7.3 shows that although the two lines are not regularly distributed, they
manifest a general trend: weather permitting/being increases and the weather per-
mitting/being decreases. The reason is that weather permitting/being has been
grammaticalized into a fixed expression, realizing the relation of condition. The
formal indicator of grammaticalization is the loss of the article the. The con-
struction the weather permitting/being has not been grammaticalized into a fixed
expression because there exists the article the in form on the one hand, and on the
other hand, it can realize different relationships. In the following, we will analyze
the trend of grammaticalization of the absolute clauses we collected from COHA
by reference to the articles the and a (Fig. 7.4).
An important type of grammaticalization of absolute clauses is concerned
with prepositions, among which the construction “noun + preposition + noun”
can be taken as a typical example pattern. Absolute clauses of attendant circum-
stance such as hat in hand are compressed semi-idiomatic absolute constructions,
and they can be expanded into augmented absolute constructions (e.g., with a hat
in his hand) (Jespersen 1949). Based on this explanation, such phrases as hand
7.4 Time 155

Table 7.5  Most frequent absolute clauses of enhancement in COHA


Weather permitting The weather permitting Weather being The weather being
1810s 0 0 0 0
1820s 0 0 0 2
1830s 1 0 0 6
1840s 0 0 0 6
1850s 2 0 0 10
1860s 0 0 0 3
1870s 3 0 0 3
1880s 0 0 0 4
1890s 0 0 0 4
1900s 1 0 0 2
1910s 2 2 0 3
1920s 2 0 0 4
1930s 5 0 0 5
1940s 9 0 0 1
1950s 4 1 0 0
1960s 2 0 0 0
1970s 4 0 0 0
1980s 3 0 1 0
1990s 7 0 0 1
2000s 11 0 0 0
Total 56 3 1 54

70
weather permitting/being the weather permitting/being
60

50

40

30

20

10

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 7.3  Historical distributions of weather permitting/being…with and without the

in hand and side by side can be referred to as idiomatic absolute constructions.


We now take such prepositions as in, on, and by popularly used in the fixed or
semi-fixed expressions as examples to discuss the grammaticalization of absolute
clauses. For a clear visual presentation, we convert the original frequencies into
the standard frequencies of per hundred million words (see Fig. 7.5).
156 7 Discussions

90 with article no article


80

70

60

50

40

30

20

10

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Fig. 7.4  Historical distribution of absolute clauses with and without article in COHA (per hundred
million words)

3500

3000

2500

2000

1500

1000

500

0
1810s 1820s 1830s 1840s 1850s 1860s 1870s 1880s 1890s 1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s
Extension 2116 1516 2497 2823 2823 3014 3028 2668 2524 2751 2568 2830 2227 2403 2587 2081 2431 2216 2602 2645

Fig. 7.5  Historical distribution of grammaticalized absolute clauses of extension (per hundred


million words)

Although the number of absolute clauses of extension is obviously increasing


along time, Fig. 7.5 shows that the historical distribution of the fixed or semi-fixed
absolute clauses of extension is generally even.

7.5 Summary

This chapter is a discussion on the research of Chap. 6, involving the distribution


of function types, the case choice, the stylistic distribution, and the diachronic
­distribution of absolute clauses. As for the functional distribution, the quantitative
research of absolute clauses of elaboration, extension, and enhancement are based
on clause complexes. Data show that there are absolute clauses used alone as sin-
gle clauses. These absolute clauses elaborate, extend, or enhance the preceding or
following clauses, but the interdependent relationship is paratactic. The indepen-
dently used absolute clauses and absolute clauses of appositive have no double
7.5 Summary 157

transitivity structures, because they both realize parataxis. Absolute clauses of


adjunct, subject, and complement have double transitivity structures. We did not
count the absolute clauses of subject and complement, for such constructions are
large in number, and are difficult and unnecessary to extract manually.
As for the case choice, there are only a few absolute clauses with personal
­pronoun subject because the subject of absolute clauses is not co-referential with
and is less dependent on that of the primary clauses. The personal pronoun sub-
ject of absolute clauses can be either nominative or accusative. When there is a
­contrast between the subjects of two absolute clauses, their case is arbitrary, but
when the antecedent does not appear in the preceding context and there is not a
contrast formed, the personal pronoun subject is always accusative. Absolute
clauses with accusative personal pronoun subject are mostly extensive. Those of
enhancement are the dependent finite clauses with conjunctions omitted. When
changed into absolute clauses, the personal pronoun subjects remain nominative.
As for the stylistic distribution, different function types of absolute clauses
are different. However, except fiction, absolute clauses tend to occur in no other
­registers. This is because a key feature of fiction is the dynamic narration of
­stories and the static description of sceneries, and absolute clauses can be used
to highlight the primary clauses. The reason why there are few absolute clauses
in academic is that language of academic texts is characterized by passive voice
and nominalization. Passive voice ensures the objectivity of argumentation and
nominalization is a method to compact information. Therefore, there are ­relatively
fewer clause complexes in academic texts, and hence having fewer absolute
clauses which are mostly elaborative.
As for the diachronic distribution, although the absolute clauses of enhance-
ment are decreasing in number, they are by no means disappearing and are
­basically stabilized in recent decades. This is mainly because several ­commonly
used absolute clauses of enhancement have been being grammaticalized into
fixed expressions. A comparative study of several commonly used absolute
clauses with article and their grammaticalized forms without article shows that
absolute clauses with article are decreasing in number, indicating that abso-
lute clauses are gradually grammaticalized. The grammaticalization of absolute
clauses can also be reflected in the semi-idiomatic and idiomatic constructions of
noun + preposition + noun.

References

Huang, G.-W. (1998). A functional analysis of the English causative structure. Journal of Foreign
Languages, 1, 12–16.
Jespersen, O. (1949). A modern English grammar on historical principles. London: Allen & Unwin.
Onions, C. T. (1905 [2010]). An advanced English syntax. Whitefish: Nabu Press.
Quirk, R., Greenbaum, S., Leech, G., & Svartvik, J. (1985). A comprehensive grammar of the
English language. London, New York: Longman.
Wheelock, F. M., & LaFleur, R. A. (2005). Wheelock’s latin (6th ed.). New York: HarperCollins.
Chapter 8
Conclusion

8.1 Main Findings of This Research

The findings of this research can be summarized by the answers to the questions
proposed in Chap. 4.
What types of relation can absolute clauses realize?
Absolute clauses as non-finite clauses with subject have the same meaning
potential as all non-finite clauses. However, not all non-finite clauses with sub-
ject can form absolute clauses. Absolute clauses defined by traditional grammar
can be classified into two syntactic types: clausal adjuncts and attendant circum-
stances. Relevant research proposes a third type: appositives. According to the
relation system of clause complex in SFL, the three types of absolute clauses cor-
respond to the dependent clauses of extension, enhancement, and elaboration in
clause complexes. Since absolute clauses can realize hypotactic expansion, they
have the potential to realize hypotactic projection. According to the principle of
double transitivity analysis, absolute clauses can be seen as embedded non-finite
clauses realizing circumstances and participants. In addition, the corpus-based
research shows that besides realizing hypotaxis, there are also a large number of
absolute clauses standing alone, i.e., they are used as independent sentences. In
spite of being independent, absolute clauses still construe a relation of elabora-
tion, extension, and enhancement with the preceding or following sentences. This
kind of relation is cohesive but not structural. Further analysis shows that abso-
lute clauses can realize not only hypotaxis but also parataxis. Four types of abso-
lute clauses can be recognized in structure: absolute appositive clauses, absolute
adjunct clauses, absolute subject clauses, and absolute complement clauses.
Different types of absolute clauses are different in independence. The analysis
of independence of absolute clauses from two dimensions, the independent ten-
dency of absolute clauses and that of primary clauses, show that different types of
absolute clauses are on a cline of independence. The independently used absolute

© Springer-Verlag Berlin Heidelberg 2015 159


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3_8
160 8 Conclusion

clauses are the most independent, absolute clauses in paratactic clause complexes
are the second most independent, and then absolute clauses in hypotactic clause
complexes, the least independent. In hypotaxis, absolute clauses of extension are
more independent than those of enhancement. Absolute of participants are the
least independent because they are restricted by the main verbs of the primary
clauses. Seen from the relation between the subjects of absolute clauses and the
main verbs, absolute clauses of subject are more independent than those of com-
plement. But seen from the requirement of the whole construction by the main
verb, absolute clauses of complement are more independent than those of subject.
The general principle is that the closer the relation between the absolute clauses
and the main verbs is, the less independent the absolute clauses are, or vice versa.
What are the synchronic and diachronic distributions of absolute clauses?
To answer this question, we extracted from the Brown Family Corpora,
the BNC, and COHA 10,930 absolute clauses of expansion and 47,889 absolute
clauses of projection. The quantitative diachronic and synchronic research on
these absolute clauses shows that the number of occurrences of absolute clauses
in the recent 200 years is not decreasing but increasing. This is mainly reflected in
the obvious increase of absolute clauses of extension. The traditional opinion that
absolute clauses are decreasing can only explain the tendency of absolute clauses
of enhancement. Corpus data do not show that absolute clauses of enhancement
are diminishing. The data of the most recent 50 years show that the distribution of
absolute clauses of enhancement tends to be stable. Analysis shows that in the pro-
cess of historical development, absolute clauses have been being grammaticalized.
Some absolute clauses of enhancement have been fixed into stereotyped expres-
sions, realizing some certain functions, hence are impossible to disappear.
The observation based on the Brown Family Corpora shows that absolute
clauses are mostly in fiction and are seldom in learned. This is obviously not in
agreement with traditional grammar according to which absolute clauses are popu-
lar in formal texts, because the register of learned is the most formal, with that of
press, general prose, and fiction following. However, the distribution of absolute
clauses in these four registers is exactly the other way round, respectively, fiction,
general prose, press, and learned. There are no spoken data in the Brown Family
Corpora. The BNC-based research shows that absolute clauses are rarely seen in
either the informal spoken texts or the formal learned texts and still occur in fic-
tions. Therefore, the distribution of absolute clauses is different in register, but this
difference has nothing to do with whether the register is formal or not. The rea-
son why absolute clauses are mostly used in fiction is that there are narrations of
events and descriptions of scenes. Absolute clauses turn out to be equipped with
such functions.
In addition, we have carried out a research based on the Brown Family Corpora
on the case choice of absolute clauses discussed in traditional grammar. Of the 61
absolute clauses with personal pronoun subject, most are concentrated in fiction
and general prose, and fiction dominates. This is mainly because the total num-
ber of words of the register of fiction is larger. When transferred into the stand-
ard frequency of per million words, the frequency in general prose is obviously
8.1  Main Findings of This Research 161

higher than that in fiction, even the average frequency in fiction is slightly lower
than the total average frequency. The study of the case distribution on the 49 case
marked absolute clauses with personal pronoun subject shows that the personal
pronoun subject can be either nominative or accusative and the distribution of case
in the recent 50 years does not show any significant diachronic or regional differ-
ences. However, such a small number of absolute clauses with personal pronoun
subject show significant register differences. In all the four registers, the number
of nominative pronouns is clearly smaller than that of accusative pronouns. The
difference between the nominative pronouns and the accusative pronouns is the
largest in general prose, and the proportion in fiction is nearly at the total aver-
age level. Only in press are there more nominative pronoun subjects than accusa-
tive pronoun subjects, and the number of nominative pronoun subjects is larger
than that in fiction. Personal pronouns do not tend to occur in the learned regis-
ter, neither do absolute clauses with personal pronoun subjects. In the six corpora,
there are only two absolute clauses with itself as subject. Data show that when the
subject of an absolute clause has to be a personal pronoun, people do not tend to
choose nominative pronoun subject, rather they tend to choose accusative pronoun
subject. The COHA-based research presents the same result. The case of personal
pronoun subjects of absolute clauses shows obvious diachronic difference in the
span of 200 years. Although the total number of nominative pronoun subjects is
much larger than that of accusative pronoun subjects, the distribution of personal
pronoun subjects shows that the number of nominative pronouns has been decreas-
ing and that of accusative pronouns increasing since the beginning of the twentieth
century. In addition, the number of accusative pronoun subjects began to surpass
that of nominative personal pronoun subjects in the 1980s phase.

8.2 Limitations and Further Research

8.2.1 Limitations

The corpus-based systemic functional research of absolute clauses has the follow-
ing problems:
First, there is no way to extract all absolute clauses from the corpus through
automatic retrieval and manual extraction.
The advantage of computer processing lies in its being able to process more
data quickly and to reveal some implicit language features. However, there is
still a considerable distance between computer processing and manual process-
ing in depth, accuracy, flexibility, richness, etc. “Automatic analysis gets harder
the higher up we move along the hierarchy of stratification” (Halliday and
Matthiessen 2004: 49). That is to say, the higher the grammatical rank is, the more
difficult the automatic analysis will be. For example, automatic analyses can deal
with any models described in words and models of lower lexico-grammatical
ranks, but cannot conduct systemic functional analyses of clauses completely or
162 8 Conclusion

conduct meaning analyses, because meaning is vague and ambiguous in nature.


Therefore, it certainly will be difficult to retrieve absolute clauses at a higher level
along the hierarchy of stratification.
In addition, the relations realized by clause complexes are meaning-based
rather than form-based. However, it is the form but not the meaning that is
retrieved from the corpus. Most of the index lines retrieved using the regular
expressions we wrote cannot form absolute clauses. Although the index lines in
line with the formal construction of absolute clauses can be retrieved by com-
puters, the extraction of absolute clauses is totally by hand. It is a hard work to
retrieve all the index lines that are possible to form absolute clauses from a corpus
with a vocabulary of hundreds of millions of words, and then to extract absolute
clauses by hand. Therefore, we have to limit the search condition. Eventually, we
extracted manually nearly 60 thousand absolute clauses of expansion and projec-
tion. However, these are still a small part of the total number of absolute clauses
actually used in the corpora.
Different forms of absolute clauses have different function orientations. For
example, absolute clauses with adverbial groups or prepositional phrases as predi-
cate usually realize extension, those having present participles especially being in
the predicate usually realize enhancement of cause, those having past participles
in the predicate tend to realize enhancement of time, and those with personal pro-
noun subjects usually realize elaboration. The regular expressions we wrote have
reduced the number of absolute clauses of extension, especially in COHA we only
extracted absolute clauses with present participles as predicates. Although the
number of absolute clauses of extension has been dominant, the number of exten-
sion may further increase if all formal types of absolute clauses are included.
Second, the identification of functional categories is inevitably subjective.
Absolute clauses are not introduced by explicit conjunctive expressions. The
identification of their relation types is inevitably subjective. Moreover, according
to SFL, there is not a clearly dividing line between categories, and the members of
two categories are on a cline. This is why many absolute clauses may have multi-
ple interpretations. However, when we identify the relation types realized by abso-
lute clauses, we can only ascribe them to the most possible interpretation.

8.2.2 Further Research

“As Jones said, ‘a science without difficulties is not a science at all’” (Halliday and
Matthiessen 2004: 35). Many scholars advocate seeking a balance “between vol-
ume of analysis and richness of analysis: low-level analysis can be automated to
handle large volumes of text, but high-level analysis has to be carried out by hand
for small samples of text” (ibid.: 49). Absolute clauses as non-finite clauses are at
a higher level in rank and so are suitable for manual processing. This requires that
in the future research, a special attention be paid to the intercrossing between con-
cepts and the complementarity of methods: complementarity between quantitative
8.2  Limitations and Further Research 163

and qualitative researches and complementarity between manual and automatic


processing. With the continuous perfection of the SFL theory and the development
of computer technology, the automatic processing will gradually develop from the
low-level analysis to high-level analysis in order to automate the retrieving and
extracting of absolute clauses to a higher and higher degree, and thus, it is possible
to extract all absolute clauses from the corpus.
With a large enough number of absolute clauses, we not only can make our
research result more convincing and persuasive, but also possibly carry out further
research on the corresponding relations between the forms and functions of abso-
lute clauses. In the present research, most absolute clauses we extract from the
corpora are composed of nominal groups and participles, especially present par-
ticiples, and the research on the function types of absolute clauses is also based
on this type of construction. Then, if all formal types of absolute clauses distin-
guished by traditional grammar are included, will the distribution of relation types
of absolute clauses change? If the answer is positive, which formal type of abso-
lute clauses has led to this change? Or in other words, do different formal types of
absolute clauses realize different types of relation? Data show that different forms
of absolute clauses have different function orientations. Most absolute clauses
with adverbial groups or prepositional phrases realize extension; absolute clauses
with participial predicate tend to realize enhancement of time. In the grammati-
calized expressions, weather/time permitting realizes enhancement of condition
and there/it being realizes enhancement of reason. These fixed expressions will to
some extent affect the functional distribution of absolute clauses. That is, if all the
absolute clauses with prepositional phrase as predicate had been included, the pro-
portion of the extension type would have increased, but on the other hand, due to
the restriction of the regular expressions, many absolute clauses such as it being…
are excluded. If this type of absolute clauses had been included, then the propor-
tion of enhancement would have increased.
Grammatical metaphor plays an important part in functional syntax. It occurs
from “a realignment between a pair of strata: a remapping of the semantics on to
the lexico-grammar” (Halliday 1978: 192) and is “the expression of a meaning
through a lexico-grammatical form which originally evolved to express a differ-
ent kind of meaning” (Thompson 1996: 165). Grammatical metaphor can occur in
all the three metafunctions. Since absolute clauses are an alternative form of finite
clauses, can they be seen as a metaphorical form of finite clauses? If so, what
types of grammatical metaphors are contained in absolute clauses?
Expansion and projection are semantic concepts. At the syntactic level, the
projected clauses can function as either complement or subject. Why are the pro-
jected clauses of complement dependent clauses in clause complexes, but those of
subject embedded in single clauses? In other words, are the complement clauses
higher in rank than the subject clauses?
Further research on these questions will not only help to understand the nature
of absolute clauses, but also enrich the syntactic theory of SFL.
164 8 Conclusion

References

Halliday, M. A. K., & Matthiessen, C. M. I. M. (2004). An Introduction to Functional Grammar


(3rd ed.). London: Edward Arnold.
Halliday, M. A. K. (1978). Language as Social Semiotic: The Social Interpretation of Language
and Meaning. London: Edward Arnold.
Thompson, G. (1996). Introducing Functional Grammar. London: Edward Arnold.
Appendix A
TreeTagger POS Tagset

CC Coordinating conjunction
CD Cardinal number
DT Determiner
EX Existential there
FW Foreign word
IN Preposition or subordinating conjunction
JJ\w* Any adjective
JJ Adjective
JJR Adjective, comparative
JJS Adjective, superlative
LS List item marker
MD Modal
N\w+ Any noun
NN Noun, singular or mass
NNS Noun, plural
NP Proper noun, singular
NPS Proper noun, plural
PDT Predeterminer
POS Possessive ending
PP Personal pronoun
PP$ Possessive pronoun
RB\w* Any adverb
RB Adverb
RBR Adverb, comparative
RBS Adverb, superlative
RP Particle
SYM Symbol
TO To
UH Interjection
VB\w* Any form of the “be” verb

© Springer-Verlag Berlin Heidelberg 2015 165


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3
166 Appendix A: TreeTagger POS Tagset

VB Verb “be” base form


VBD Verb “be” past tense
VBG Verb “be” gerund or present participle
VBN Verb “be” past participle
VBP Verb “be” present tense, other than third person singular
VBZ Verb “be” third person singular present
VH\w* Any form of the “have” verb
VH Verb “have” base form (have)
VHD Verb “have” past tense (had)
VHG Verb “have” gerund or present participle (having)
VHN Verb “have” past participle (had)
VHP Verb “have” present tense, other than third person singular
VHZ Verb “have” third person singular present (has)
VV\w* Any form of any lexical verb
VV Lexical verbbase form
VVD Lexical verbpast tense
VVG Lexical verbgerund or present participle
VVN Lexical verbpast participle
VVP Lexical verbpresent tense, other than third person singular
VVZ Lexical verbthird person singular present
W[\w\$]+ Any wh-word
WDT Wh-determiner
WP Wh-pronoun
WP$ Possessive wh-pronoun
WRB Wh-adverb
\s Space
\S+_[A-Z0-9\$]+ Any word_tag
\S+_\S+ Anything_tag
Appendix B
Tables of Corpus Data Statistics

See Tables B.1, B.2, B.3, B.4, B.5, B.6, B.7, B.8 and B.9.

Table B.1  Regional distribution of absolute clauses in the Brown Family Corpora


British American
Feature Percent N Percent N T Stat Significance ChiSqu Significance
Function N = 182 N = 200
Elaboration 18.1 33 29.5 59 2.61 +++ 6.74 +++
Extension 63.2 115 61.0 122 0.44 0.19
Enhancement 18.7 34 9.5 19 2.61 +++ 6.72 +++
Style N = 182 N = 200
Press 8.2 15 7.0 14 0.46 0.21
Prose 20.9 38 21.5 43 0.15 0.02
Learned 4.4 8 6.0 12 0.70 0.49
Fiction 66.5 121 65.5 131 0.20 0.04
British N = 182 N = 0
Lob 31.9 58 0.0 0 0.00 0.00
Flob 30.2 55 0.0 0 0.00 0.00
Clob 37.9 69 0.0 0 0.00 0.00
American N = 0 N = 200
Brown 0.0 0 34.5 69 0.00 0.00
Frown 0.0 0 31.0 62 0.00 0.00
Crown 0.0 0 34.5 69 0.00 0.00
+ Weak significance (90 %), ++ Medium significance (95 %), +++ High significance (98 %)

© Springer-Verlag Berlin Heidelberg 2015 167


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3
168 Appendix B: Tables of Corpus Data Statistics

Table B.2  Stylistic distribution of absolute clauses in the Brown Family Corpora


Press Prose Learned Fiction
Feature Percent N Percent N Percent N Percent N
Function N = 29 N = 81 N = 20 N = 252
Elaboration 20.7 6 42.0 34 45.0 9 17.1 43
Extension 48.3 14 39.5 32 20.0 4 74.2 187
Enhancement 31.0 9 18.5 15 35.0 7 8.7 22
Region N = 29 N = 81 N = 20 N = 252
British 51.7 15 46.9 38 40.0 8 48.0 121
American 48.3 14 53.1 43 60.0 12 52.0 131
British N = 15 N = 38 N = 8 N = 121
Lob 33.3 5 60.5 23 25.0 2 23.1 28
Flob 53.3 8 21.1 8 62.5 5 28.1 34
Clob 13.3 2 18.4 7 12.5 1 48.8 59
American N = 14 N = 43 N = 12 N = 131
Brown 50.0 7 32.6 14 25.0 3 34.4 45
Frown 42.9 6 39.5 17 16.7 2 28.2 37
Crown 7.1 1 27.9 12 58.3 7 37.4 49

Table B.3  Relation distribution of absolute clauses in the Brown Family Corpora


Elaboration Extension Enhancement
Feature Percent N Percent N Percent N
Style N = 92 N = 237 N = 53
Press 6.5 6 5.9 14 17.0 9
Prose 37.0 34 13.5 32 28.3 15
Learned 9.8 9 1.7 4 13.2 7
Fiction 46.7 43 78.9 187 41.5 22
Region N = 92 N = 237 N = 53
British 35.9 33 48.5 115 64.2 34
American 64.1 59 51.5 122 35.8 19
British N = 33 N = 115 N = 34
Lob 51.5 17 25.2 29 35.3 12
Flob 36.4 12 27.8 32 32.4 11
Clob 12.1 4 47.0 54 32.4 11
American N = 59 N = 122 N = 19
Brown 27.1 16 36.1 44 47.4 9
Frown 40.7 24 25.4 31 36.8 7
Crown 32.2 19 38.5 47 15.8 3
Appendix B: Tables of Corpus Data Statistics 169

Table B.4  Relation distribution of absolute clauses in BNC


Elaboration Extension Enhancement
Feature Percent N Percent N Percent N
Style N = 321 N = 1,500 N = 114
Spoken 1.9 6 1.8 27 2.6 3
Newspapers 0.3 1 2.4 36 7.0 8
Magazine 7.5 24 5.9 89 14.0 16
Misc 14.3 46 12.4 186 22.8 26
Academic 10.0 32 2.4 36 6.1 7
Nonacademic 15.0 48 6.1 92 14.9 17
Fiction 51.1 164 68.9 1,034 32.5 37
Form N = 321 N = 1,500 N = 114
-ing 41.4 133 38.4 576 51.8 59
-en 16.2 52 20.1 302 14.0 16
adj 14.0 45 12.7 190 7.0 8
prep 28.3 91 28.8 432 27.2 31

Table B.5  Form distribution of absolute clauses in BNC


-ing -en adj prep
Feature Percent N Percent N Percent N Percent N
Style N = 872 N = 397 N = 340 N = 602
Spoken 1.5 13 2.5 10 2.6 9 1.3 8
Newspapers 2.6 23 2.0 8 2.4 8 3.0 18
Magazine 6.5 57 4.8 19 5.0 17 9.8 59
Misc 15.4 134 12.1 48 28.5 97 11.6 70
Academic 2.8 24 3.0 12 2.9 10 4.8 29
Nonacademic 7.3 64 7.6 30 7.1 24 9.1 55
Fiction 63.9 557 68.0 270 51.5 175 60.3 363
Function N = 768 N = 370 N = 243 N = 554
Elaboration 17.3 133 14.1 52 18.5 45 16.4 91
Extension 75.0 576 81.6 302 78.2 190 78.0 432
Enhancement 7.7 59 4.3 16 3.3 8 5.6 31
170

Table B.6  Stylistic distribution of absolute clauses in BNC


Spoken Newspapers Magazine Misc Academic Nonacademic Fiction
Feature Percent N Percent N Percent N Percent N Percent N Percent N Percent N
Form N = 40 N = 57 N = 152 N = 349 N = 75 N = 173 N = 1,365
-ing 32.5 13 40.4 23 37.5 57 38.4 134 32.0 24 37.0 64 40.8 557
-en 25.0 10 14.0 8 12.5 19 13.8 48 16.0 12 17.3 30 19.8 270
-adj 22.5 9 14.0 8 11.2 17 27.8 97 13.3 10 13.9 24 12.8 175
-prep 20.0 8 31.6 18 38.8 59 20.1 70 38.7 29 31.8 55 26.6 363
Function N = 36 N = 45 N = 129 N = 258 N = 75 N = 157 N = 1,235
Elaboration 16.7 6 2.2 1 18.6 24 17.8 46 42.7 32 30.6 48 13.3 164
Extension 75.0 27 80.0 36 69.0 89 72.1 186 48.0 36 58.6 92 83.7 1,034
Enhancement 8.3 3 17.8 8 12.4 16 10.1 26 9.3 7 10.8 17 3.0 37
Appendix B: Tables of Corpus Data Statistics
Appendix B: Tables of Corpus Data Statistics 171

Table B.7  Relation distribution of absolute clauses in COHA


Elaboration Extension Enhancement Independent
Feature Percent N Percent N Percent N Percent N
Time N = 1,319 N = 5,137 N = 1,132 N = 1,828
1810 0.8 11 0.1 4 0.3 3 1.5 27
1820 0.8 11 0.4 19 3.9 44 0.4 8
1830 3.4 45 1.2 62 8.8 100 1.5 27
1840 2.7 36 1.8 94 6.5 74 2.3 42
1850 4.6 61 2.4 124 11.0 125 2.9 53
1860 4.0 53 2.4 121 7.8 88 6.1 111
1870 2.7 36 2.4 125 9.6 109 6.9 126
1880 4.5 59 2.8 146 5.7 64 2.1 39
1890 2.7 35 3.4 174 8.9 101 2.4 43
1900 2.2 29 3.6 186 6.6 75 3.0 55
1910 3.2 42 3.2 165 5.1 58 6.0 110
1920 3.6 48 5.8 300 5.6 63 5.5 100
1930 4.4 58 6.4 329 4.1 46 4.6 84
1940 6.6 87 6.0 308 3.4 38 3.4 63
1950 6.8 90 6.3 325 1.9 21 4.7 86
1960 9.8 129 7.8 402 2.3 26 7.1 130
1970 6.2 82 8.4 433 2.3 26 10.8 198
1980 8.3 110 8.5 438 1.6 18 7.9 144
1990 11.4 151 11.9 610 2.6 29 10.4 191
2000 11.1 146 15.0 772 2.1 24 10.4 191
Subject N = 1,319 N = 5,137 N = 1,132 N = 1,828
Noun 69.7 919 94.0 4,827 88.0 996 89.8 1,642
Pronoun 30.3 400 6.0 310 12.0 136 10.2 186
Noun N = 919 N = 4,827 N = 996 N = 1,642
No-article 60.3 554 61.8 2,984 29.6 295 75.6 1,242
Article 39.7 365 38.2 1,843 70.4 701 24.4 400
Pronoun N = 400 N = 310 N = 136 N = 186
General 71.8 287 41.6 129 47.8 65 30.1 56
Personal 28.2 113 58.4 181 52.2 71 69.9 130
Case N = 113 N = 181 N = 71 N = 130
Subjective 63.7 72 53.0 96 49.3 35 41.5 54
Objective 6.2 7 24.9 45 8.5 6 10.8 14
No-case 30.1 34 22.1 40 42.3 30 47.7 62
172 Appendix B: Tables of Corpus Data Statistics

Table B.8  Noun and pronoun subject distribution of absolute clauses in COHA


Noun Pronoun
Feature Percent N Percent N T Stat Significance ChiSqu Significance
Time N = 8,384 N = 1,032
1810 0.5 40 0.5 5 0.03 0.00
1820 0.9 75 0.7 7 0.71 0.50
1830 2.1 179 5.3 55 6.23 +++ 38.69 +++
1840 2.7 227 1.8 19 1.65 2.71 +
1850 3.5 297 6.4 66 4.50 +++ 20.18 +++
1860 3.7 309 6.2 64 3.91 +++ 15.29 +++
1870 4.5 377 1.8 19 4.01 +++ 16.08 +++
1880 2.9 240 6.6 68 6.36 +++ 40.33 +++
1890 3.4 286 6.5 67 4.92 +++ 24.17 +++
1900 3.9 327 1.7 18 3.48 +++ 12.10 +++
1910 4.1 341 3.3 34 1.20 1.43
1920 5.0 417 9.1 94 5.54 +++ 30.61 +++
1930 5.5 462 5.3 55 0.24 0.06
1940 5.4 452 4.3 44 1.53 2.34
1950 5.5 459 6.1 63 0.83 0.70
1960 7.1 595 8.9 92 2.12 ++ 4.49 ++
1970 8.0 673 6.4 66 1.84 + 3.38 +
1980 7.7 648 6.0 62 1.98 ++ 3.90 ++
1990 10.9 915 6.4 66 4.49 +++ 20.10 +++
2000 12.7 1,065 6.6 68 5.71 +++ 32.45 +++
Noun-type N = 8,384 N = 0
No-article 60.5 5,075 0.0 0 0.00 0.00
Article 39.5 3,309 0.0 0 0.00 0.00
Pronoun-type N = 0 N = 1,032
General 0.0 0 52.0 537 0.00 0.00
Personal 0.0 0 48.0 495 0.00 0.00
Case N = 0 N = 495
Subjective 0.0 0 51.9 257 0.00 0.00
Objective 0.0 0 14.5 72 0.00 0.00
No-case 0.0 0 33.5 166 0.00 0.00
Function N = 8,384 N = 1,032
Elaboration 11.0 919 38.8 400 25.07 +++ 589.48 +++
Extension 57.6 4,827 30.0 310 17.02 +++ 281.01 +++
Enhancement 11.9 996 13.2 136 1.21 1.46
Independent 19.6 1,642 18.0 186 1.20 1.43
+ Weak significance (90 %), ++ Medium significance (95 %), +++ High significance (98 %)
Appendix B: Tables of Corpus Data Statistics 173

Table B.9  Case distribution of personal pronoun subjects of absolute clauses in COHA


Subjective Objective
Feature Percent N Percent N T Stat Significance ChiSqu Significance
Time N = 257 N = 72
1810 1.2 3 0.0 0 0.00 0.85
1820 0.4 1 0.0 0 0.00 0.28
1830 6.2 16 1.4 1 1.64 2.69
1840 0.0 0 0.0 0 0.00 0.00
1850 8.2 21 2.8 2 1.59 2.52
1860 3.9 10 2.8 2 0.44 0.20
1870 2.3 6 0.0 0 0.00 1.71
1880 6.6 17 1.4 1 1.73 + 2.97 +
1890 9.3 24 1.4 1 2.26 ++ 5.06 ++
1900 1.6 4 0.0 0 0.00 1.13
1910 2.7 7 2.8 2 0.02 0.00
1920 11.7 30 12.5 9 0.19 0.04
1930 5.8 15 9.7 7 1.17 1.36
1940 4.7 12 4.2 3 0.18 0.03
1950 3.9 10 4.2 3 0.11 0.01
1960 16.3 42 9.7 7 1.39 1.94
1970 4.7 12 6.9 5 0.77 0.59
1980 7.4 19 6.9 5 0.13 0.02
1990 1.6 4 19.4 14 6.22 +++ 34.80 +++
2000 1.6 4 13.9 10 4.72 +++ 21.00 +++
Function N = 257 N = 72
Elaboration 28.0 72 9.7 7 3.25 +++ 10.32 +++
Extension 37.4 96 62.5 45 3.89 +++ 14.52 +++
Enhancement 13.6 35 8.3 6 1.20 1.44
Independent 21.0 54 19.4 14 0.29 0.08
+ Weak significance (90 %), ++ Medium significance (95 %), +++ High significance (98 %)
Appendix C
Corpus Retrieving Demonstration

See Figs. C.1, C.2 and C.3.

© Springer-Verlag Berlin Heidelberg 2015 175


Q. He and B. Yang, Absolute Clauses in English from the Systemic Functional
Perspective, The M.A.K. Halliday Library Functional Linguistics Series,
DOI 10.1007/978-3-662-46367-3
176 Appendix C: Corpus Retrieving Demonstration

Fig. C.1  Example retrievals of noun or pronoun + participle constructions from the LOB


Corpus using the regular expression: \S+_(,|SENT|:)\s(\S+_(DT|CD|P\w*)\s)*(\S+_JJ\w*\s)
*\S+ _(PP|N\w+ )\s(\S+ _RB\w*\s)*\S+ _V(B|D|H|V)[GN]\s(\S+ _IN\s)*
(\S+_(DT|CD|P\w*)\s)*(\S+_(PP|N\w+)\s)*\S+_(,|SENT|:)
Appendix C: Corpus Retrieving Demonstration 177

Fig. C.2  Example retrievals of personal pronoun + nonfinite element constructions from the Frown
Corpus using the regular expression: \S+_(,|SENT|:)\s\S+_PP(\S+_RB\w*)*\s(\S+_TO\s\
S+_V[BDHV]|\S+_(V(B|D|H|V)[GN]|JJ\w*|IN|DT|N\w+)
178 Appendix C: Corpus Retrieving Demonstration

Fig. C.3  Example retrievals of noun or pronoun + present participle constructions from BNC
using the regular expression: [y*] [nn*]|[p*] [v?g*] (0-2) [y*]