Review of The Acquisition of Complex Sentences
Date: Fri, 23 Sep 2005 00:20:24 -0700 (PDT)
From: Jason Brown
Subject: The Acquisition of Complex Sentences
AUTHOR: Diessel, Holger
TITLE: The Acquisition of Complex Sentences
SERIES: Cambridge Studies in Linguistics 105
PUBLISHER: Cambridge University Press
Jason Brown, Department of Linguistics, University of British Columbia
This book is a study of the acquisition of complex sentences in
English. The book consists of eight chapters, plus a substantial
appendix. This book will be of interest to researchers working on
language acquisition, and syntax in general. It would be well suited
for use in a graduate course in language acquisition, and portions
would even be suitable for undergraduate courses in the same field of
Chapter 1, Introduction: In this chapter, an outline of the structure of
the book is presented, and what constitutes a "complex sentence" is
discussed. This includes sentences with a matrix and a subordinate
clause, or coordinate clauses. Furthermore, three different types of
subordinate clause are distinguished: complement, relative, and
adverbial clauses, all three of which can be finite or nonfinite. A
complement clause serves as the argument of the predicate in a
superordinate clause, a relative clause as an attribute of N or NP, and
an adverbial clause serves as a modifier of the matrix clause or VP.
The data used in this study is then discussed. The data consists of
12,000 multiple clause utterances from 5 English speaking children
obtained from the CHILDES database. As the author notes, this is to
date the largest database compiled for this type of study. The two
hypotheses behind the study are then laid out. The first is that
complex sentences evolve from simple sentences. The author states
that "the development of complex sentences originates from simple
nonembedded sentences that are gradually 'transformed' to multiple-
clause constructions" (3). The second hypothesis is that "children's
early complex sentences are organized around concrete lexical
expressions. More schematic representations of complex sentences
emerge only later when children have learned a sufficient number of
lexically specific constructions to generalize across them" (3).
Chapter 2, A dynamic network model of grammatical constructions:
This chapter lays out the theoretical background of the study. In
contrast to a generative approach, the author takes a "functional-
cognitive approach". The study makes use of both Construction
Grammar (Lakoff 1987, Fillmore 1988, Fillmore & Kay 1993, Goldberg
1995, Croft 2001, etc.) and the usage-based approach to grammar
(Langacker 1988, 2000, Bybee 1995, 2001; also Barlow and Kemmer
2000). The chapter first gives a brief overview of Construction
Grammar. Construction Grammar views constructions as grammatical
primitives. In this way, grammatical constructions are like words in that
they are pairings of sound and meaning. The chapter discusses the
importance of formulaic "pre-fabricated chunks", "low-level formulas"
and idiomatic expressions to the theory. Also discussed is the
continuum between the grammar and lexicon inherent in this view.
The chapter next discusses the usage-based approach, which views
grammar as being shaped by usage. The major differences between
the usage-based model and the generative model are discussed, and
they include views about innateness and grammatical development.
Chapter 3, Towards a definition of complex sentences and
subordinate clauses: The goal of this chapter is to define the
phenomena at hand, specifically as "grammatical constructions that
express a specific relationship between two (or more) situations in two
(or more) clauses" (41), where the terms "situation" and "clause" are
given further elaboration. Most of the chapter is dedicated to defining
subordinate clauses. Subsections describe the syntactic and
semantic features, as well as the processing of subordinate clauses.
The chapter concludes with a summary of the features of prototypical
Chapter 4, Infinitival and participial complement constructions: This
chapter deals with the first complex sentences to appear in the
acquisition data: non-finite complement constructions. After a
literature review on the subject, an overview of non-finite complement
constructions in adult English is provided. A claim of this chapter is
that the earliest examples in acquisition of this type of construction
can actually be analyzed as a type of quasi-modal (with no
embedding). The next forms that appear are the non-finite
complement clauses that are truly embedded. Of these constructions,
the earlier forms are typically subject control, followed by object
control. This path of development is attributed to clause expansion.
Diessel shows that while the earlier constructions constituted
propositions that made reference to a single situation, the complement
clause and complement-taking verb in later constructions that emerge
are less tightly bound to each other and can be considered two
propositions. Finally, several motivations for the order of acquisition
are discussed. The frequency of constructions in the ambient
language is pointed out as one factor, while the complexity of
constructions is another.
Chapter 5, Complement clauses: This chapter deals with the next
complex sentence type to appear in the acquisition data, which is the
finite complement clause. One claim is that early complement clauses
are accompanied by formulaic matrix clauses; that is, only a single
proposition is expressed, and thus, complement clauses at this stage
are not embedded. The chapter argues that the matrix clauses in this
stage are non-assertive (i.e. they are epistemic markers, attention
getters, etc.). The clauses are then divided into 4 different types:
1) epistemic markers,
2) deontic modality markers,
3) discourse directives, and
4) say, tell, and pretend.
The chapter also discusses performative vs. assertive uses, and the
role that pragmatics and illocutionary force play. The general course
of development for complement clauses is also outlined, whereby the
formulaic constructions occur, then the performative, then the
assertive. This is shown to be another case of clause expansion,
whereby a single proposition expands in later development into two
independent propositions. Again, factors motivating this development
can be found in the ambient language, in the complexity of the
constructions, and also in the cognitive capacities of the child.
Chapter 6, Relative clauses: In this chapter, Diessel discusses the
emergence of relative clauses. In a review of the literature, some of
the more relevant hypotheses concerning relative clauses in
acquisition are discussed, such as the noninterruption hypothesis, the
filler-gap hypothesis, the NVN-schema hypothesis, the parallel-
function hypothesis, and the conjoined-clause hypothesis. Also
discussed are restrictive vs. non-restrictive relative clauses, and the
conjoined clause analysis. The chapter outlines the development of
relative clauses from simple, lexically-specific constructions into fully
fledged bi-clausal structures. The order of development is outlined.
The first relatives to emerge are described as predicate nominal
amalgams in which the relative clause is not independent of the matrix
clause. The next constructions to emerge are fully bi-clausal
structures in which two independent propositions are expressed.
Much like the infinitival and complement clauses discussed in the
previous two chapters, the acquisition of relative clauses involves an
incremental development from simple to more complex clauses. This
development is again of the "clause expansion" type that was seen for
non-finite and finite complement clauses.
Chapter 7, Adverbial and co-ordinate clauses: This chapter discusses
how these two types of clause are traditionally distinguished
(adverbial clauses are typically classified as subordinate clauses,
coordinate clauses are typically considered as non-embedded), and
how they actually form a continuum (the class of these two clauses is
termed "conjoined clauses"). The contrast between these clauses
and those discussed in chapters 4-6 is their path of development.
While non-finite and finite complement clauses and relative clauses
develop through clause expansion, adverbial and coordinate clauses
develop through the integration of two independent clauses. As
Diessel states, "conjoined clauses functioning as independent
sentences emerge prior to conjoined clauses that are integrated in
biclausal constructions" (170). Finally, factors influencing the order of
acquisition of these clauses are discussed, including frequency in the
ambient language, complexity of processing, and discourse-pragmatic
Chapter 8, Conclusion: This chapter provides an overview of the
study, as well as an in-depth discussion of the hypotheses laid out in
the introduction and the relevance of the data from chapters 4 through
7. The chapter discusses the various complex sentence types, and
how their development relates to the usage based model and
Cognitive Grammar. Special emphasis is placed on how the paths of
development for each of the complex sentence types in the book
relates to the two hypotheses of the study; in particular, how complex
sentences start as simple clauses, and how they emerge as lexically
specific constructions and develop into constructional schemas. The
appendix at the end of the book (pp. 186-199) provides 14 pages of
The book is extremely readable, even for an audience that may not be
familiar with the syntax of complex sentences or syntactic acquisition.
For example, in chapters 4-7, Diessel follows a consistent formula for
chapter breakdowns, which include a literature review, followed by an
outline of the construction in adult grammar, an exposition of the
acquisition data, an analysis, and finally, a discussion. Readers
should find this format extremely accessible for two reasons: first, it
makes each chapter easy to digest, and second, it makes each
chapter directly relatable to the others.
At first glance, one might criticize Diessel for basing an acquisition
study solely on English data. However, given the scarcity of data
available from other languages, and the abundance of English data, it
made for a strong empirical grounding for the study. The use of five
subjects, along with the fairly wide age range of the subjects, gave the
study an admirable amount of breadth. Diessel must be applauded for
providing the first comprehensive look at ALL complex sentence types
in acquisition, and also for the use of data in the study. Diessel
consistently uses data from the same 5 subjects, and the overall
utterance counts are impressive. Such a study should set a standard
for other researchers, especially those interested in how complex
sentences are acquired by speakers of languages other than English.
For researchers working on syntactic acquisition in other languages,
this book should ultimately leave you wondering, would clause
expansion and integration look the same in these other languages?
Hopefully Diessel's work will stimulate others to seek an answer to this
Although it's an excellent overview of Cognitive Grammar and the
usage-based model, upon first encounter, one may wonder what
chapter 2 is doing in the book. The necessity of the chapter becomes
evident in the following chapters and in the conclusion, where the
discussion of the data forces the reader to seriously think about
concepts like formulaicity and pre-fabricated chunks, as well as all of
the factors that help to determine paths of development, such as the
frequency of constructions in the ambient language, the complexity of
constructions, and the cognitive and processing demands that certain
constructions require. Diessel makes it a point to show how all of
these factors are relevant for the acquisition of complex sentences.
Finally, the hypotheses are exciting. The idea that all complex
sentences emerge from simple sentences seems simple enough, but
the claim that complex sentences arise through two different paths
(expansion AND integration) can only be taken seriously with loads of
empirical data to support it. Diessel provides exactly this.
The book is of the same high quality that characterizes the other
books in the Cambridge Studies in Linguistics series. The exciting
proposals and empirical rigour of the study make this both an
interesting and well-grounded work. The book should become a
regular reading in graduate courses in the acquisition of syntax, and it
should set an example for future studies to follow.
Barlow, Michael & Susanne Kemmer (eds.) (2000). Usage-Based
Models of Language. Stanford: CSLI Publications.
Bybee, Joan (1995). Regular morphology and the lexicon. Language
and Cognitive Processes 10:425-455.
Bybee, Joan (2001). Phonology and Language Use. Cambridge:
Cambridge University Press.
Croft, William (2001). Radical Construction Grammar. Oxford: Oxford
Fillmore, Charles J. (1988). Mechanisms of construction grammar.
Berkeley Linguistics Society 14:35-55.
Fillmore, Charles J. & Paul Kay (1993). Construction Grammar.
Berkeley: University of California Press.
Goldberg, Adele E. (1995). A Construction Grammar Approach to
Argument Structure. Chicago: University of Chicago Press.
Lakoff, George (1987). Women, Fire, and Dangerous Things.
Chicago: Chicago University Press.
Langacker, Ronald W. (1988). A usage-based model. In B. Rudzka-
Ostyn (ed.), Topics in Cognitive Linguistics, pp. 127-161. Amsterdam:
Langacker, Ronald W. (2000). A dynamic usage-based model. In
Barlow & Kemmer (2000), pp. 24-63.
| ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Jason Brown is currently a Ph.D. student at the University of British
Columbia. His research focus is on phonological theory, with special
interests in the phonetics-phonology interface, phonological
representations, and feature theory. He is also interested in the
acquisition of syntax.