AUTHOR: Brinton, Laurel J.
TITLE: The Comment Clause in English
SUBTITLE: Syntactic Origins and Pragmatic Development
SERIES: Studies in English Language
PUBLISHER: Cambridge University Press
Daniela Kolbe, Department of English Studies, University of Trier
''The Comment Clause in English'' is a comprehensive study of the developments
and functions of clausal pragmatic markers in English. It is aimed at scholars
interested in diachronic and syntactic studies. Most centrally, it focuses on
the question of whether or not comment clauses are instances of grammaticalization.
Chapters 1-3 offer a detailed and informative overview of comment clauses and
related topics. They discuss the most important terms and concepts for the
following study by giving an extensive review of previous research. Chapters 4
to 10 contain case studies of the individual derivation of seven different types
of comment clauses, such as I MEAN or those with SEE, e.g., LET'S SEE, YOU SEE.
Brinton's overall qualitative analysis is complemented by comparisons of the
frequency of the different types of comment clauses across time periods, text
types, and topical areas. She illustrates her investigation with a large number
of corpus-based examples.
Chapter 1 ''Introduction: comment clause, parentheticals, and pragmatic
markers'' provides an introduction to comment clauses and related constructions,
such as sentence adverbials, disjunct adverbials, parentheticals, and pragmatic
markers. It concludes by giving an overview of the study's approach, sources,
Chapter 2 ''Semantic and syntactic development of pragmatic markers'' then
discusses how pragmatic markers can develop in general, providing an overview of
existing literature on the topic. Brinton first deals with the semantic
development from referential meaning to non-referential meaning with discourse
functions. The second, larger part of the chapter is devoted to the description
of different syntactic paths in the derivation of pragmatic markers.
Chapter 3 defines and reviews the ''Processes of change'' that have proved
relevant in previous research for the development of pragmatic markers in
general and comment clauses in particular. These are grammaticalization,
pragmaticalization, lexicalization, idiomaticization, and
(inter)subjectification. Grammaticalization is presented as the most central
Chapters 4-10 present case studies and are similar in their structure. First,
the Present-Day English use of each type of comment clause is discussed, taking
into account frequencies in different text styles as well as regional or
diachronic variation. After this introduction, the historical development of
each type of comment clause is examined. Supported by corpus evidence, the
semantic as well as the syntactic development of the respective comment clause
are described. In the conclusion of each case study, Brinton assesses if the
development of the respective comment clauses can be described within the
framework of grammaticalization.
The case studies deal with the following kinds of comment clauses: ''clauses
with SAY'' (chapter 4), I MEAN (chapter 5), ''clauses with SEE'' (chapter 6),
''IF YOU WILL and AS IT WERE'' (chapter 7), ''clauses with LOOK'' (chapter 8),
''WHAT'S MORE and WHAT ELSE'' (chapter 9), ''epistemic/evidential parentheticals
- I GATHER and I FIND'' (chapter 10). Thus, chapter 5 is the only case study
that examines only one type of comment clause. In the case studies that deal
with comment clauses containing the same verb, Brinton distinguishes between
different types of these clauses (e.g. LET ME SEE vs. SEE vs. YOU SEE) and their
subtypes. These types and subtypes do not only depend on the form and use of the
relevant comment clause in Present-Day English, but also on their different
In the conclusion in chapter 11, Brinton reviews the theoretical background of
the study and unites the results of the case studies in a description of the
development of comment clauses in general. She points out in particular the
shortcomings of the ''matrix-clause'' hypothesis by Thompson and Mulac (1991a
and b). Thompson and Mulac stated that the deletion of the complementizer THAT
in sentences such as I THINK THAT IT'S GOOD > I THINK IT'S GOOD leads to a less
tight connection between I THINK and IT'S GOOD. Thus, as I THINK loses its
matrix clause status, the complement clause becomes the main clause, and I THINK
may become a comment clause as in IT'S GOOD, I THINK.
''The Comment Clause in English'' is an extensive and detailed diachronic study
of the subject in question. It thus complements the existing research on
pragmatic markers. Both the theoretical overviews (of previous research,
linguistic terminology, and processes of change) in the first chapters and the
analyses in the following chapters are detailed and comprehensive. The lists of
examples from Old to Present-Day English are illuminating and attest to the
amount of text work that underlies this study. Brinton always provides examples
from all the periods of English in which these constructions existed, which
yields a lively illustration of the development of each construction.
Brinton's study contains thorough overviews of the previous literature and
explanations of the central terms for the study. She reviews different accounts,
theories, definitions, and sources, including frequent direct quotes from these
sources. Overall, Brinton provides comprehensible analyses and conclusions,
e.g., when she pinpoints the difference between literal and cognitive SEE (p.
148) or the difference between the development of YOU SEE vs. I SEE (p. 155),
which is particularly well described.
Although the outlines of all case studies look similar, they are variable enough
to allow for the individual history and function of each discussed item. The
organization of each chapter leads to the derivation that is proposed. The
analyses are carefully conducted and easy to understand. No proposed derivation
In sum, it is an impressive and excellent study of the derivation of comment
clauses. Hence, any flaws or disadvantages can only be found in the following
minor aspects and details.
It surprised me that the comment clause I THINK did not receive more attention
in this book. Certainly, it has already attracted much attention in linguistic
research. This is mostly due to the fact that I THINK is the prototypical
example used for the ''matrix clause hypothesis'' (Thompson and Mulac 1991 a and
b). In fact, one central motivation for Brinton's case studies was to test
whether comment clauses actually arise from an increase in the deletion of THAT,
as suggested by the matrix clause hypothesis (p. 241). However, the historical
development of I THINK is never elaborated on. According to Palander-Colin
(1997), I THINK develops in parallel to METHINKS, and thus Brinton's examples
contain only METHINKS and not I THINK (p. 38). The immediately following
example, PRITHEE, is discussed in a lot more detailed, with several examples
(pp. 38-39). Since Brinton points out that the matrix clause hypothesis is not
always supported by the historical data (pp. 39-40, 247), I had expected a
renewed, diachronic investigation of I THINK. The (existence of the) author's
own diverging hypothesis on the development of comment clauses with matrix
clause structure is only revealed at the end of the theoretical discussion of
the matrix clause hypothesis (pp. 44).
In her case studies, Brinton tests the matrix clause hypothesis for all comment
clauses with a potential matrix clause structure (e.g., I (DARE)SAY, I FIND, I
GATHER, I/YOU SEE). She points out already in the introduction (p. 14) that
THAT-less clauses were the norm in Old English (see Rissanen 1991).
Consequently, the historical evidence for the development suggested by the
matrix clause hypothesis is often shaky. This testing of the matrix clause
hypothesis is one of the greatest benefits of this book. Brinton shows that what
looks so straightforward from a synchronic perspective may in fact be a lot more
complicated when one considers diachronic linguistic evidence. The matrix clause
claim is indeed difficult to verify if the use of THAT was so rare in earlier
stages of English. Thompson and Mulac's analysis is based on Present-Day English
conversations (1991a: 315). In several cases, Brinton provides a more plausible
derivation: She derives I MEAN from I MEAN +phrasal structure (Brinton 2008:
124-127), YOU SEE from AS YOU SEE (pp. 154-157) and I FIND from AS I FIND (pp.
235-237). However, in the discussion of the origin of (I) SAY (pp. 90/110) and I
DARESAY (pp. 97/110) she presents no alternative to the matrix clause
hypothesis, even though the data do not support it.
Brinton's case studies consist of careful analyses, and the sheer number of the
forms she analysed is impressive. Only in two minor points does her analysis not
seem indisputable. Firstly, Brinton establishes a fine-grained distinction of
different forms of SAY with 6 subtypes (pp. 74-80), but it is not very clear how
some of these subtypes relate to each other. Subtypes 5 and 6 have two further
subtypes: 5a, 5b and 6a, 6b. It seems that the a and b types of 5 and 6 are
differentiated by more features than they have in common. SAY 5 a and b have in
common that they both belong to the category of quasi-interjections and that
they are pragmatic markers, but they fulfill different functions. SAY 6 a and b,
however, are not in the same category, nor are they both pragmatic markers. They
also do not have the same function, or the same historical development. SAY 6b
shares one of the listed features (being a pragmatic marker) with SAY 5 (a/b),
but none with SAY 6a. Thus it seems that the types 5/6 a/b could also be
classified as types 5/6/7/8.
Secondly, among all the detailed discussions of the derivations of all comment
clauses in the study, there is one which does not seem completely plausible:
LOOKIT. Brinton rejects the derivation of LOOKIT < LOOK AT (p. 198) for the
non-literal use of LOOKIT as comment clause. However, it remains unclear whether
the derivation of LOOKIT from LOOKAT has ever been stated for the non-literal
use of LOOKIT. According to Brinton, the hypothesis that LOOKIT derives from
LOOKAT is based on the following quotation in the OED: ''LOOK-AT used among
school children for look! ... Cf. LOOK-IT in Mass. Mich.'' This quotation,
however, appears to refer to LOOKIT in its literal meaning, for which Brinton
confirms that ''[i]t seems fairly clear that [it] derived from LOOK AT'' (p.
198). She argues that LOOKIT derives from LOOK TO'T, a clause type that is
widely attested in the corpus data. However, none of the provided examples show
a spelling that suggests a beginning erosion of the phonetic development from
LOOK TO'T to LOOKIT. The necessary phonetic changes in the vowel sound from /U/
to /I/ and the elision of /t/ are not discussed in the analysis of the development
of LOOKIT. Hence, this derivation remains partly unexplained, though not
implausible. The phonetic changes necessary for this derivation are only
mentioned in the chapter's concluding overview of features of grammaticalization
that apply for all LOOK forms (p. 201).
Although the overall aim of Brinton's study is a qualitative analysis, she
presents and compares frequencies and ratios of different forms, in Modern
English as well as in different diachronic, altogether 18, corpora. Sometimes,
however, striking differences in the charts are not discussed in the
accompanying text. The ratios between the non-parenthetical to parenthetical use
of AS IT WERE (figure 7.3) are 27:37 in the Middle English part of the Helsinki
Corpus but 16:2 in the Canterbury Tales (p. 174). The ratios of the three
functions of WHAT ELSE (pragmatic function: literal main clause interrogative:
literal finite relative clause) are 57:19:18 in the British National Corpus
(BNC), and 3:10:8 in the Strathy corpus (figure 9.3). Thus in the British (BNC)
data WHAT ELSE predominantly occurs in its pragmatic function (57:37), whereas
the Canadian (Strathy) data contain mostly literal uses (18:3) of WHAT ELSE (p.
212-213). Both findings are not mentioned in the description of the respective
charts. A few pages later, however, the differences between the BNC and the
Strathy corpus are taken to show differences between British and Canadian usage
in general (figure 10.1, p. 224). Perhaps Brinton does not comment on the first
two findings because they draw on a relatively low number of occurrences.
Another interesting finding that is stated, but not commented on, is the low
frequency of WHAT'S MORE in oral genres, and its wider use in written
non-fiction, including academic, writing (p. 204). As comment clauses typically
occur more often in oral genres, I would have been interested in some
elaboration on this point, e.g., why it is exactly this comment clause that
Despite its overall dense structure, the study also contains sections which do
not seem absolutely necessary. The distinction between comment clauses and
declarative matrix clauses with first (pp. 37-41) vs. second/third-person
subject (p. 41) seems redundant, since the section on second or third person
subjects is very short and does not provide any new insights. The comparisons
with other verbs in chapters 4 and 8 do not provide much information, although
they are interesting. The discussion of the fact that HARK(EN), LISTEN and HEAR
have similar pragmatic uses as LOOK (pp. 199-200) seems a bit too detailed for
the additional information gained from it. This is also true for the comparison
of SAY-comment clauses with LIKE and WHAT in Present-Day English in section 4.3
(pp. 80-82), which only states that there are similarities in usage. The
information in each of these sections is valuable, but the question is whether
the individual sections are necessary.
Another minor detail that could be improved upon concerns the lists of comment
clauses in chapter 4, which are not always consistent. Sometimes, (I) SAY and
SAY appear as individual items on the list, sometimes the list only contains (I)
SAY. Hence, the five comment clauses with SAY including (I) SAY and (AS) YOU SAY
listed in an enumeration (p. 73) can represent seven possible types, when I SAY
is differentiated from SAY and AS YOU SAY from YOU SAY. They are, however,
illustrated by only six examples (pp. 73-74). Brinton also varies between using
only THAT IS TO SAY, e.g. (p. 73), and sometimes THAT IS (TO SAY), e.g. (p. 74).
This makes it difficult to determine the exact meaning of the parentheses and
the difference between SAY (< I SAY) or just SAY.
The fact that the outline is the last part of the introduction (section 1.7.3,
pp. 21-23) informs readers that ''chapter 1 conceptualises forms such as I KNOW
and YOU SEE ...'' at the end of precisely this chapter, which seems slightly
odd. It might be more reader-friendly to acknowledge that the reader has read
most of the first chapter already or to omit the description of the introduction
from the outline and to focus on the remainder of the book instead.
Something that contributes greatly to reader-friendliness are visualizations of
findings, in particular the itemised list that specifies how which
characteristics of grammaticalization apply to the development of LOOK-forms
(p.201), the overview of the development of the different types of LOOK clauses
(p.202), and the date charts in chapter 9 and in the conclusion that show when
which function of which clause type begins and ends to be used (figure 9.4, p.
217, figure 11.2, p. 248). However, the fact that they do appear at all makes
one notice that they do not appear in the other case studies. Had I not seen
them in chapter 9, I would not have missed them elsewhere. Providing more visual
representations of the central results throughout would have fitted in nicely
with the overall similar structure of the case studies. Nevertheless, the
overview in figure 11.2 (p. 248) is one of the best features of the book. It
does not only list all kinds of comment clauses discussed in the preceding
study, including expressions that may have changed into others over time, such
as AS YOU SEE > SEE (adding up to an impressive number of 36), it also sorts
them according to four structurally different categories of their sources (such
as imperative verb vs. adverbial / relative clause) and shows during which
periods from Middle to Present-Day English and in which function these comment
clauses were used.
I would also like to point out some slips of the pen that might distract or
slightly confuse the reader. The gravest one occurs when Quirk et al.'s
observation that YOU SEE serves to claim ''hearer's attention'' (1985: 1115) is
misquoted as ''claiming speaker's attention'' (Brinton 2008: 135). In the
development of I DARESAY, I am not sure if it really ought to be ''matrix I DARE
THAT'' and parenthetical ''I DARE'' (p. 95) or rather I DARE ''SAY'' (THAT). The
description of imperative IF YOU WILL as '''if you willing to do so''' (p. 178)
seems to lack an ARE as in 'if you ARE willing to do so.' The paraphrase of I
GATHER as ''I think know that it is like this'' (p. 226) seems to mean ''I think
I know that it is like this'' ╨ unless the use of the serial-verb construction
THINK KNOW was intended. One list of supposedly ''five pragmatic meanings of I
MEAN'' contains only four items: (a), (b), (c), and (e)[sic!] (114). On page 141
it says ''(11d) presents an interesting example in which SEE is used in both
ways in the same sentence,'' although there is no (11d). (12d), however,
contains the respective instances of SEE and it is (12a)-(12c) that illustrate
the functions of SEE? and not, as stated, (11a-c).
Although Brinton uses comprehensive terminology, some terms are used
inconsistently. This may cause confusion when they refer to different linguistic
units. On page 207, the reader finds ''clause internal'', ''clause external'' in
figure 9.2, but ''sentence internal'', ''sentence external'' in the accompanying
text. The predominantly used term ''matrix clause'' changes to ''main clause''
in section 10.4, e.g., in the phrase ''main clause hypothesis'', and the quote
continues ''that is, that parenthetical I FIND originates in a main clause
structure with a clausal THAT-complement'' (p. 236), confirming that ''main
clause'' is used synonymously with ''matrix clause'' here. On page 237, ''matrix
clause hypothesis'' occurs again. It also seems unusual that in the discussion
of the syntactic development of comment clauses, adverbial and (nominal)
relative clauses (pp. 43-47) appear as subtypes of matrix clauses (2.3.3).
Page 61 contains a concise definition of pragmatic markers: ''The fact that
pragmatic markers do not belong to a readily identifiable word class, typically
occupy an extra sentential position, have non-truth-conditional meaning, and
function pragmatically suggests that they may not be part of 'grammar proper'
and hence do not result from a process of grammaticalization.'' This definition
would have also been perfect in the introduction section on pragmatic markers
(pp. 14-18) insofar as it pinpoints the need to introduce the different
processes of change.
After all this nitpicking, I would like to emphasize another major quality of
Brinton's study. First of all, the controversial points that are relevant for
all case studies are presented and critically assessed in the introductory
theoretical chapters. This acquaints readers early enough with the central
issues important for the remainder of the book. These three chapters contain all
the information the reader needs to in order to fully grasp the discussions in
the case studies even if he or she is not an expert in pragmatic markers,
grammaticalization, or diachronic linguistics in general. All examples from Old
and Middle English are translated into present-Day English.
In sum, I can recommend this book to all linguists interested in discourse
markers, grammaticalization, and historical syntax and semantics.
''look, n.'' The Oxford English Dictionary. 2nd ed. 1989. OED Online. Oxford
University Press. 22 July 2009 <http://dictionary.oed.com/cgi/entry/50135443>
''lookit, int. and v.'' The Oxford English Dictionary. 2nd ed. 1989. OED Online.
Oxford University Press. 20 July 2009
Gelderen, Elly van (2006) A history of the English Language.
Amsterdam/Philadelphia: John Benjamins.
Kemenade, Ans van (1994) Old and Middle English. In: Eckehard König, Jan van der
Auwera. The Germanic Languages. London: Routledge, 110-141.
Millward, Celia M. (1989) A biography of the English Language. Fort Worth: Holt,
Rinehart and Winston.
Palander-Collin, Minna (1997) A medieval case of grammaticalization, methinks.
In Matti Rissanen, Merja KytÜ and Kirsi Heikkonen, eds. Grammaticalization at
work, Berlin / New York: Mouton de Gruyter, 371-403.
Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech and Jan Svartvik (1985) A
comprehensive grammar of the English language. London: Longman.
Rissanen, Matti (1991) On the history of that/zero as object clause links in
English. In Karin Aijmer and Bengt Altenberg, eds. English corpus linguistics:
Studies in honour of Jan Svartvik, London / New York: Longman, 272-289.
Thompson, Sandra and Anthony Mulac (1991a) A quantitative perspective on the
grammaticization of epistemic parentheticals in English. In Elizabeth Closs
Traugott and Bernd Heine, eds. Approaches to grammaticalization, volume 2,
Amsterdam: John Benjamins, 313-329.
Thompson, Sandra and Anthony Mulac (1991b) The discourse conditions for the use
of the complementizer 'that' in conversational English. Journal of Pragmatics
ABOUT THE REVIEWER
Daniela Kolbe is assistant professor of English linguistics in the Department of
English studies at the University of Trier. Her research interest lies mainly in
syntactic variation -- especially in dialects. Her teaching interests also
include the history of the English language and conversation analysis.