Review of Metadiscourse in L1 and L2 English
AUTHOR: Ädel, Annelie
TITLE: Metadiscourse in L1 and L2 English.
SERIES: Studies in Corpus Linguistics 24
PUBLISHER: John Benjamins
Louisa Buckingham, Sabanci University; Universidad de Granada.
This book builds upon the rich existing work on metadiscourse theory and
practice, and, by exploring the application of computer assisted methods to
studies of metadiscourse, it contributes to both our theoretical and
practical appreciation of metadiscourse through empirical research on its
use by British and US students and Swedish learners of English. Through
studying its use in argumentative texts (a genre thought to make generous
use of metadiscourse) by Swedish learners of English, and two groups of
native speakers (British and American), the author identifies how
metadiscourse use varies across British and American English argumentative
writing, and how it is used differently in advanced learner writing. All
essays are derived from the International Corpus of Learner English (ICLE)
at the University of Louvain, Belgium.
The acceleration of studies in metadiscourse over the last decades is
possibly partly motivated by the broad issue of developing academic English
competence of NNS for international publication. Foreign language fluency
does not translate into the ability to manage metadiscourse, and its
inappropriate use in writing leads to perceptions of the text being too
dense, difficult to follow, or, alternatively, informal and convoluted.
Interestingly, learner metadiscourse is not characterized by error, but
rather inappropriateness: underuse or misuse. In light of this, corpus
analysis appears the most satisfactory means of examining metadiscourse
use, as it is the cumulative effect more than individual occurrences that
characterize the use of metadiscourse in texts.
Consisting of seven main chapters (plus a conclusion and appendices), the
author divides the theoretical background between the early chapters and
the final chapter (and the appendices), which enables the reader to move
quickly into the main research. The main contributions presented in
chapters 2 to 5 comprise discussions on the identity of metadiscourse, and
corpus analyses of frequency and textual distribution.
The introduction encompasses definitions of metadiscourse followed by an
outline of research methods, indicating points of convergence and
divergence with other studies. Metadiscourse seems to constitute a language
universal in both written and spoken domains, though there appears to be
little agreement in the literature of its nature or of its genre-specific
manifestations, and few studies have investigated it from a
The second chapter presents a model of metadiscourse based on Jakobson's
(1998) functional model of language, unlike previous research based on the
Hallidayan approach. In her classification of metadiscourse, the author
distinguishes between personal (i.e. the use of personal pronouns and nouns
referring to the writer or reader) and impersonal types of metadiscourse
(i.e. the use of passives and impersonal constructions). She identifies the
different orientations that it may have: text oriented metadiscourse refers
to the current text and its language use ('in this essay', 'in the
following'); writer-oriented metadiscourse refers to the writer persona
('finally, I would like to discuss the topic'; 'as I stated above');
reader-oriented metadiscourse refers to the imagined reader of the text
('you may be thinking', 'so you see', 'there were many reasons for'); and
finally, participant-oriented metadiscourse, a mixture of the two previous
categories ('What do we mean by...then?'; 'as we have seen'; 'therefore I
will give you some sort of background that might have some relevance for
how you picture...').
Ädel's conceptualization of metadiscourse differs from certain previous
studies as she excludes stance markers (expressing, for instance,
uncertainty, disagreement, disbelief: 'I am quite convinced', 'I think it
is of great importance', 'I am in favour of'), as well as numerous
instances of personal pronouns, due to their reference to the 'real world'
as opposed to the text itself. Personal pronouns particularly require
careful analysis of context to ascertain whether the reader-writer
interaction expressed fulfils a metadiscoursive function, as many instances
refer to personal experiences outside the text.
As a general point of departure, the features of metadiscourse are
described in the following terms: fuzzy category membership (it is
difficult to make categorical distinctions between what is and what is not
metadiscourse); functional categories (metadiscourse can be represented
morphosyntactically by a range of different forms and structures);
multifunctionality (metadiscourse expressions may fill two or more
discourse functions simultaneously); and context dependency (context is
necessary to identify metadiscourse).
Chapters 3-5 present the results of the investigation into the use of
metadiscourse by learners and NS of English; chapters 3 and 4 examine
personal and impersonal expressions, while chapter 5 analyses the textual
distribution of these two categories.
Personal metadiscourse makes direct reference to the writer or reader
through pronouns or nouns. Specific examples are first retrieved from the
corpora and then sifted through manually to determine whether they possess
text internal or text external references. The study is both quantitative
(as the frequency of use among the three language communities is recorded),
and qualitative as the study examines the discourse functions of personal
metadiscourse among the three language communities).
Quantifying personal metadiscourse involves decisions regarding the unit of
measurement. As the author notes, different approaches have been taken; for
example, Mauranen (1993) uses the sentence as the unit of analysis. Ädel
maintains this would not enable a true quantitative analysis, as learners
tend to cluster units of personal metadiscourse within one sentence;
therefore, to verify this tendency, each unit needs to be counted as one
The results demonstrate that, within this particular corpus, considerable
difference exists across language communities with regard to the frequency
and density of use of metadiscourse. Swedish students use more than twice
that of US students, while these use twice as much as British students.
Metadiscourse is also more densely clustered in learner texts.
The author subsequently turns to analyzing metadiscourse qualitatively by
identifying the discourse functions it serves. Initially considering
existing taxonomies by Vassileva (1998), and Kuo (1998), Ädel then
formulates her own compendium of 16 functions, distinguishing between two
subcategories: metatext and reader-writer interaction. The former includes
functions such as: saying, defining, exemplifying, reminding, adding,
arguing, introducing topics, focusing, concluding, contextualizing; while
the latter comprises functions that anticipate the reader's reaction such
as: clarifying, aligning perspectives, imagining scenarios, hypothesizing
about the reader, and appealing to reader. Calculating the occurrence of
these functions across the three language communities, the Swedish learners
again out-perform others in the frequency of use of each of these functions.
Corpus searches of impersonal metadiscourse in chapter 4 focus on a series
of search terms (such as: essay, word, conclu*, second(ly), begin*,
question, answer) compiled by reading the essays in the corpus and by
consulting the existing literature. Impersonal metadiscourse can also be
classified in terms of discourse functions, and Ädel identifies four main
categories: phoric markers (cataphoric or anaphoric; known together as
endophoric markers in Hyland's 1998 terminology), references to text/code
(for example: text, paper, essay, section, in other words), code glosses,
discourse labels (explicitly announcing discourse acts such as defining,
summarizing, concluding). Similar to the results obtained for personal
metadiscourse, Swedes writing in English heavily overuse these features.
Chapter 5 examines the textual distribution of both impersonal and personal
metadiscourse with a view to testing the initial hypothesis across the
three language communities that these language features typically appear at
beginnings and endings of text. To retrieve data, the author employs the
concordance programme Wordsmithtools (Scott 1999), which is able to divide
a text into distinct parts and quantify occurrences of search terms in
each. The initial hypothesis is verified across the three groups.
Where the previous chapters have dealt with whether and how the use of
metadiscourse varies, chapter 6 considers the possible causes of this
variation, considering, for example, genre comparability, cultural
conventions, and learner strategies. With regard to genre, the author
discerns that metadiscourse is likely to find increased use where greater
concern exists for how readers understand and accept ideas posited in the
text (hence, argumentative writing seems to be a particularly propitious
genre). An examination of cultural conventions would require a corpus
analysis of L1 Swedish writing in order to study how this differs from
native Anglophone texts (for example, to determine whether Swedish written
discourse constitutes a more reader responsible or writer responsible
culture, to use Hinds's 1987 terminology). As such a corpus is nonexistent,
the author settles for a description of a general tendency towards
informality in Swedish spoken and written language use, and claims that
foreign language teaching in Sweden favours the spoken language. This leads
to the claim that the overuse of metadiscourse may be a result of lack of
Finally, with regard to learner strategies, the author discusses the
possibility that learners find that greater explicitness of writer presence
and discourse acts facilitates communication in a foreign language. It
seems reasonable to suggest that, as writing formally in a foreign language
requires considerable cognitive effort, explicit use of metadiscourse helps
the writer manage the task of constructing a text.
Chapter 7 returns to a more abstract level in its description of theories
of metadiscourse, motivated by the perceived need to establish greater
theoretical rigour. The author classifies previous work as either
representing broad or narrow approaches to metadiscourse studies, situating
her own work between the two.
In the final chapter, besides summarizing the main theoretical and
empirical findings, the author proposes topics for further research such
as: the study of essays in L1 Swedish, a comparison of metadiscourse use
among English learners with different L1s, the comparison of learner data
with professional writing (as opposed to L1 student writing), and a
phraseological study on metadiscourse in terms of constructions,
prefabricated phrases, and idiomatic/conventionalized expressions.
The book closes with a collection of appendices comprising short
discussions on topics related to the book's central theme which serve as
optional additional reading. These include: comparability of corpora; the
control corpus and the norm; metadiscourse as non-propositional material;
meta-terminology: terms to talk about metadiscourse and related phenomena,
how they differ or interrelate.
This is a study which is of high interest both to people working in
different linguistic fields and to educators working in the area of L2
language and writing instruction. The appropriate use of metadiscourse, as
this study clearly demonstrates, plays a key role in achieving the
appropriate degree of writer visibility, as well as appropriate modes of
interaction with the reader, both of which contribute to the approximation
of native speaker communicative competence in formal written discourse.
Those involved in advanced language teaching will gain important insights
into what to incorporate into teaching curricula, as well as how to deal
with clumsy sounding texts due to overloaded use of metadiscourse.
Ädel has explored different ways of displaying qualitative and quantitative
information, using a variety of graphs, tables and concordance lines, as
well as textual excerpts, which, chosen prudently, permit a clear overview
of the main tendencies explored in her discussion.
Finally, her objective discussions of the results obtained from the corpus
include an evaluation of how the nature of the corpus may favour certain
results. For example, unlike the English NS writers, the fact that the
Swedish learners of English composed their texts under examination
conditions without recourse to other texts could well have influenced their
overuse of metadiscourse for strategic reasons, for example, to disguise
their difficulties in fulfilling the required essay length. The importance
of the book's content and the methods employed, matched by the author's
fluid, engaging writing style (displaying a masterful command of
metadiscourse herself), makes the book an absorbing, satisfying read.
Hinds J. (1987). Reader versus writer responsibility: A new typology. In:
U. Conner & R. Kaplan (Eds.), Writing across languages: Analysis of L2
text. Reading, MA: Addison-Wesley Publishing Company.
Hyland, K. (1998). Exploring corporate rhetoric: Metadiscourse in the
CEO's letter. Journal of Business Communication, 35(2), 224-245.
Jakobson, R. (1998). On language: Roman Jakobson. L.R. Waugh & M.
Monville-Burston (Eds.). Cambridge, MA and London: Harvard University Press.
Mauranen, A. (1993) Cultural differences in academic rhetoric A
textlinguistic study. Frankfurt am Main: Peter Lang.
Vassileva, I. (1998). Who am I/we in academic writing?: A contrastive
analysis of authorial presence in English, German, French, Russian, and
Bulgarian. International Journal of Applied Linguistics, 8(2): 163-190.
Kuo, Chih-Hua (1998). The use of personal pronouns: Role relationships in
scientific journal articles. English for Specific Purposes, 18(2), 121-138.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Louisa Buckingham is currently completing her Ph.D. at the University of
Granada (Spain) in the area of phraseology, and simultaneously works as an
academic writing instructor at Sabanci University. She has published in the
areas of phraseology, and second language writing.