VIETNAM NATIONAL UNIVERSITY – HANOI
COLLEGE OF FOREIGN LANGUAGE
_______________________ PHAM THI THANH THUY HEDGING DEVICES IN ENGLISH AND
VIETNAMESE ECONOMIC RESEARCH
ARTICLES (ERAs.)
(PHƯƠNG TIỆN RÀO ĐÓN TRONG CÁC BÀI BÁO NGHIÊN CỨU
KINH TẾ TIẾNG ANH VÀ TIẾNG VIỆT)
PH.D. THESIS
English Language


English Language

Supervisors:
1. Assoc. Prof. Dr. Le Hung Tien
2. Assoc. Prof. Nguyen Quang

HANOI, 2008
I

Table of Contents
TABLE OF CONTENTS I
LIST OF TABLES III
LIST OF FIGURES V
ABBREVIATIONS AND CONVENTIONS VI
CHAPTER 1. INTRODUCTION ERROR! BOOKMARK NOT DEFINED.
1.1. RATIONALE OF THE STUDY ERROR! BOOKMARK NOT DEFINED.
1.2. OBJECTIVES OF THE STUDY AND RESEARCH QUESTIONS ERROR! BOOKMARK
NOT DEFINED.
1.3. THE CONTRIBUTIONS OF THE STUDY ERROR! BOOKMARK NOT DEFINED.
1.4. RESEARCH METHODOLOGY ERROR! BOOKMARK NOT DEFINED.
1.4.1. Data Collections Error! Bookmark not defined.
1.4.2. Procedures for the Analysis of the CorpusError! Bookmark not defined.
CHAPTER 2. REVIEW OF THE LITERATURE - 22 -
2.1. EXPLORING THE CONCEPT OF HEDGING 22 -

4.2.1. Content-oriented Functions of Hedging in the Vietnamese Corpus 154
4.2.2. Reader-oriented Functions of Hedging Devices in the Vietnamese
Corpus 161
CHAPTER 5. COMPARING HEDGING USAGES BETWEEN ENGLISH AND
VIETNAMESE WRITERS 170
5.1. GENERAL COMPARISON AND CONTRAST 170
5.2. COMPARISON AND CONTRAST IN DIFFERENT CATEGORIES 172
5.2.1. Devices in the Two Corpora with No Differences 175
5.2.2. Differences among Hedging Devices in the Two Corpora 179
CHAPTER 6. CONCLUDING REMARKS AND IMPLICATIONS 186
6.1. CONCLUDING REMARKS 186
6.2. IMPLICATIONS OF THE STUDY 190
6.2.1. Implications for Language Awareness 190
6.2.2. Implications for research 191
6.2.3. Implications for language learning 192
6.3. SUGGESTED FOR FURTHER STUDY 194
III

List Of Tables
Table 1-1 Feature Matrix of Data Base Sections Error! Bookmark not defined.
Table 2.1 Hedging taxonomy (Salager-Meyer, 1994:56) - 45 -
Table 3.1 Relative Numbers of Hedging Devices per 100 words in RAs on Economics,
Applied linguistics and Physics - 58 -
Table 3.2 Descriptive Statistics: Economics, Applied Linguistics, Physics - 59 -
Table 3.3 One-way ANOVA: English Economics, Applied Linguistics and Physics- 60 -
Table 3.4 Categories of Lexical Devices in the English Corpus - 61 -
Table 3.5 Distribution of Modal Auxiliaries in the English Corpus - 66 -

linguistics, and Vietnamese Sociology 167
Table 4.15: One-way ANOVA Vietnamese Economics, Vietnamese Applied Linguistics
and Vietnamese Sociology 169
Table 5.1. Number of Hedging Devices per 100 words in Vietnamese and English
Economic Discourses 170
Table 5.2 Descriptive Statistics of Hedging Devices in the English and Vietnamese
corpus 171
Table 5.3 ANOVA Hedges in English and Vietnamese Economics 173
IV

Table 5.4 ANOVA of Phrasal Hedging Devices in the English and Vietnamese
Economics 176
Table 5.5 ANOVA of English and Vietnamese Lexical Verbs as Hedges 177
Table 5.6 ANOVA of English and Vietnamese Adverbs as Hedges 178
Table 5.7. ANOVA of English and Vietnamese Adjective as Hedges 178
Table 5.8 ANOVA of English and Vietnamese Modalities as Hedges 180
Table 5.9 ANOVA of English and Vietnamese Nouns/Pronouns as Hedges 182
Table 5.10 ANOVA of English and Vietnamese Compound Hedges 182
Table 5.11 ANOVA of English and Vietnamese Non-lexical Devices as Hedges 183

V

List Of Figures
Figure 3.1 String Matching of CAN in the English Corpus - 68 -
Figure 3.2 String Matching of COULD, COULD BE in the English Corpus - 69 -
Figure 3.3. String Matching MAY in the English Corpus - 71 -
Figure 3.4 String Matching WILL in the English Corpus - 74 -
Figure 3.5 String Matching of SUGGEST in the English Corpus - 79 -
Figure 3.6 String Matching of APPEAR and SEEM in the English Corpus - 82 -
Figure 3.7 Strings Matching of SOME in the English Corpus - 86 -

ERAs Economic Research Articles
ES. English Sociology Texts
ESP English for Specific Purposes
FTA Face Threatening Act
H
0
Null Hypothesis
H
1
Alternative Hypothesis
IE Institute of Economics
JCR Journal of Citation Reports
JEL The Journal of Economic Literature
JEP The Journal of Economic Perspectives
JIEEP
The Journal of International Economic and Economic
Policy
N Number of occurrence
P Relative Power
R Rating/Raking of imposition
RA Research article
VII

RAs Research articles
RePEc Research Papers in Economic
Sis Specialist Informants
VA Vietnamese Applied Linguistics Texts
VE. Vietnamese Economic Texts
VS. Vietnamese Sociology Texts


1. To identify the hedging devices used in the discussion sections of the English
and the Vietnamese ERAs.
2. To identify the communicative functions of hedges in Discussion sections of
English and Vietnamese ERAs.
3. To compare/contrast hedging expressions in Discussion sections in English
ERAs and in Vietnamese ERAs.

- 2 -

To realize the above objectives, this study will address the following research
questions:
1. Which typical lexico-grammatical forms are used to realize hedging functions
in English and Vietnamese ERAs.?
2. What are pragmatic functions of hedging devices in the English and
Vietnamese ERAs?
3. What are differences or similarities in the use of hedging between the two
languages?
1.3. The Contributions of the Study
Unlike previous studies focusing on hedging in primarily spoken conversation,
where it is extremely common and represents a significant communicative
resource for speakers (for example, Stubbs, 1986; Coates, 1987; Nittono, 2003;
Nguyen Quoc Sinh, 2004; Phan Thi Phuong Dung, 2004), this thesis places the
concept within academic strategies for modifying illocutionary force based on an
analysis of authentic written articles.
The research describes the distribution of surface forms used to hedge in ERAs in
English and Vietnamese. The thesis also identifies a valid explanatory framework
underlying rhetorical choices.
The analysis of these rhetorical principles provides a new area of insight for
linguists.
Knowledge about hedging from this thesis provides insights into how researchers

research papers selected from journals published in the United States and in
Vietnam. The sample collected exemplifies the typical language of Economics
and is linguistically recognizable.
The sample collected for the present analysis seems to be authentic and not
simulated since it is extracted from well known journals edited by distinguished
professors from famous universities in the USA such as Harvard University, the
University of Texas, New York University, the National Economics University
in Hanoi and Ho Chi Minh City in Vietnam. Many prominent economists publish
their real experience and case studies in these journals including Elhanan
Helpman (Harvard), Torsten Persson (London School of Economics), Eric
Rasmusen (Indiana), Robert J. Shiller (Yale), Blanchard (MIT), Le Trong Khoi,
Phan Dinh Huong (Hanoi), Phan Trong Tue (Ho Chi Minh City)… Moreover,
writers of these journals are usually aware that their audience is international and
that they have responsibility for almost all of their statements in the articles.
However, only the authenticity may not be enough because publication can have
an influence on the readership that the corporate writers are aware of and
consciously initiate. The second criterion for collecting the data is set up.
1.4.1.1.2 Reputation
Reputation has to do with the extent to which peers or the readership as a whole
rate the articles released by a given source as well as the extent to which this
source is renowned in the Economic field. Bley-Vroman and Selinkers state that
- 4 -

reputation means the extent to which the texts collected are valued by the
readership.
The sources from which data for the present analysis has been collected are
reputable sources in the Economic field. For one thing, they all have world-wide
coverage. For another, these articles are ranked by the Econometric Journal; the
RePEc (Research Papers in Economics) published by the Department of
Economics of University of Connecticut. The Econometric Journal ranks

- 5 -

sources for the purpose of generalization. Furthermore, taking care to vary the
source of material to be collected and selected may help the researcher to avoid
the pitfall of drawing conclusions from a small and poor corpus.
1.4.1.2 Criteria for Collecting the Corpus
Three criteria for the selection of the corpus for the present analysis were
identified: Universality, Text Length and Representativity.
1.4.1.2.1 Universality
The reason for choosing empirical research articles for the corpus is because of
their universality in the journal data. We have shown and summarized in table 3
the different types of research papers in the journal data. Because we are seeking
generalizations across a set of similar texts, a decision had to be made to select
for analysis not only what could be manually manageable, (i.e. not excessively
long), but especially those in particular type of research papers which were
present everywhere at once in the sample from all four sources. Moreover, there
is a hypothesis that in discussion section- one of 4 sections in an empirical
research, the writer will use a lot of hedging devices to give a series of points
referring back to statements made in the Introduction and interpret what has been
learned in the study.
This criterion and hypothesis led us to consider mostly texts from the discussion
section which was found to occur universally, as expounded in table I-1.
Type of research articles
Sample
Theory
papers
Reviews
Responses
and
reactions


- 6 -

Economic Studies

Economic & Development
Review (EDR)
- + - + +
Table 1-1 Feature Matrix of Data Base Sections
The similarity and recurrence of titles in discussion section from different sample
data are further elements that add to the criterion of the universality of material
selected.
1.4.1.2.2 Text Length
Text length has been a major criterion in deciding which text should go in the
corpus and which should not. Text length in this corpus is various because of the
content. The shortest texts chosen are 190 words (EE. Texts 9) and 354 words
(VE. Text 6); the longest are 1417 words (EE. Text 13) and 2064 words (VE.
Text 2), respectively below and above which limits a text was excluded from
consideration. Such a selection of short and fairly long texts has a certain
purpose.
1.4.1.2.3 Representative
Like the Authenticity criterion, the criterion of Representative deals with the
genuine exemplification expressed in the target model of language use.
Representative also ties with the extent to which conclusions drawn from the
analysis of a corpus are applicable to texts which may be of the same genre but
which have not been included in the corpus. The contents in the texts setting up
the corpus are real (narrowed to economic policy), life projections and up-to-data
at the moment of data collection (from the year 2002 to 2005).
A corpus of nearly 50,000 words of the present thesis is quite manageable
manually (due to the unavailability of a sufficiently large and suitable

been confirmedly kept throughout the history of the Association. At present, over
50% of the AEA membership is associated with academic institutions, 15% with
business and industry, and the remainder largely with federal, state and local
government or other non-profit organizations.
The Journals of the Association include The American Economic Review (AER),
The Journal of Economic Literature (JEL) and The Journal of Economic
Perspectives (JEP).
• Source 2: Springer
There are three things that shape Springer Science+Business Media’s business
activities: Knowledge, information and quality. With the world’s best academics
and authors (more than 150 Nobel prize-winners have published with Springer up
to the present), Springer’s publications are considered authoritative works in may
fields, read by academics and students, used by libraries and universities,
academic professionals and practitioners in various branches of industry.
Springer’s main publishing fields are economics, science, medicine, engineering,
architecture, construction and transport. In the economic field, Springer publishes
many journals, such as International Economic and Economic Policy, Economic
of Planning, Economic Theory, Empirical Economics, Experimental Economics.
- 8 -

• Source 3: The Berkeley Electronic Press (BEPress)
Launching its first journals in December of 2000. As word of the B.E.Press
model has spread, interest within academia has grown significantly. BEPress
produces tools to improve scholarly communication. These tools are innovative
and effective means to provide content production and dissemination.
The high quality peer-reviewed journals, working papers, and institutional
repository materials published by BEPress provide scholar work for the academic
community all over the world.
According to Ms. Mary Folster- Senior Academic Librarian- working at the
Memorial Library, University of Wisconsin-Madison, if we base on the impact

economy, economic theories and models in the region and other countries.
Research results are disseminated through books, monographs, articles, journals
and active programmes of conferences, seminars and workshops.
The IE publishes two periodical scientific reviews: Nghien cuu kinh te
(Economic Studies Review) (monthly in Vietnamese) and Vietnam's Socio-
Economic Development (quarterly in English).
• Source 5: Ministry of Investment -Central Institute for Economic
Management (CIEM)
Established in 1978, The Central Institute of Economic Management belongs to
the Ministry of Investment. The CIEM functions to study and propose economic
regulation and policies, economic mechanism, business environment, economic
reform. The CIEM also focuses on educating and training management officers,
consulting economic issues.
Among 95 employees of the CIEM, there are 2 associate professors, 15 doctors,
32 people hold master degrees.
Website: http://www.ciem.org.vn
• Source 6: The National Economics University (NEU)
Established more than 50 years ago, the National Economics University is a high
standard university specializing in training students at graduate and post-graduate
level in economic and business management. Besides, the NEU also has
functions to consult macro policies to the Government.
The number of lecturers and administrators of the NEU is 1117. Among them
there are 26 professors, 69 associate professors, 207 doctors, 250 masters, 20
advanced teachers, 230 senior teachers, 329 lecturers.
The NEU has a monthly journal: Economics and Development Review
1.4.1.3.2 Database
The database in this project refers to the set of publications from which articles
for analysis have been extracted. It consists of the following journals:
• The American Economic Review (AER)
This journal is published quarterly. Each of the regular issues (March, June,

economic, environmental economic, or other related fields.
This B.E. Journal in Economic Analysis and Policy produces specialist economic
and business publications that make advances in economic analysis and policy. It
also publishes articles suitable for publication in a top field journal or an
excellent general interest journal.
• Advances in Macroeconomics
In the same family as other journals such as Frontiers in Macroeconomics,
Advances in Economic Analysis & Policy, this Advances in Macroeconomics (or
The B.E. Journals in Macroeconomics) focuses on both theoretical and applied
macroeconomics. Articles in this paper are edited by distinguished professors
from Berkeley, Princeton, New York University, Boston College, Frankfurt, and
Oslo. This journal is assigned to cover papers from the broad research area
concerning modern macroeconomics. The paper meets the demand of researchers
- 11 -

in exploring such issues as unemployment, inflation, poverty, growth, and
globalization.
For the purpose of this study, each section is taken to be an independent article.
All articles taken from discussion sections in this journal have been considered
for analysis. The tables and graphs in these articles have been ignored because of
lack of sufficient verbal information.
• The Journal of International Economic and Economic Policy (JIEEP)
As a member of Springer, the Journal of International Economic and Economic
Policy publishes empirical and theoretical contributions, especially papers
relevant to economic policy. The main focus of the journal is on comparative
economic policy, international political economy, including international
organizations and policy cooperation, monetary and real/technological dynamics
in open economies, globalization and regional integration, trade, migration,
international investment, internet commerce and regulation. JIEEP aims to create
a forum for exchanging knowledge among the academic and policy community.

the journal can be accessible on the website of the HoChiMinh City University of
Economics and the Financial Times Group.
For the purpose of the present analysis, each section extracted from the
discussion sections (although they have different names) is treated as an
independent article in its own right. The reason for this decision is that, on
conceptual grounds, the sections deal with different economic issues concerning
economic policy, which relates to possibly quite different topics. Although
authors might give their discussion about the mentioned issues in other sections
of the article, the subsequent sections do not appear to have a common beginning
or introduction, nor do they appear to have a common ending or conclusion.
Therefore, each section in the article which, in this study, is the discussion
section is treated as an ARTICLE.
However, not all the articles found in this sample have been considered for
analysis for many reasons stated later in the thesis.
1.4.1.4 Extracted Research Articles
As mentioned before, discussion sections devoted to each article are taken as
independent articles. A total of 50 economic articles (25 articles in English and
25 articles in Vietnamese) have been extracted from the sample publications
mentioned previously with a view to analyzing hedging devices in their
expressions and their functions in semantic, syntactic and pragmatic aspects.
Moreover, in order to consider the use of hedging devices in different fields, 10
other articles in applied linguistics and physics genres (in English), 5 other
articles in applied linguistics and 7 articles in sociology (in Vietnamese) are also
used.
The recent issues (2000-2007) of the above journals are deliberately selected and
only empirical research articles (RAs) are chosen due to the fact that the
language used in these journals is not only authentic but updated as well. The
empirical RA is a recognizable genre and has emerged as a result of a mutually
understood communicative purpose, reporting new experiments intended to
- 13 -

are native English professors, associate professors, lecturers, journalists and
experts in economics, students in linguistics field and in other majors than
linguistics. The objectives of Vietnamese survey questionnaire are native
Vietnamese economists, university lecturers, businessmen, students in M.A.
courses in economics in National Economics University. These M.A. students are
special in various specific majors in economics and are rich in practical
experience in economic fields.
- 14 -

1.4.2. Procedures for the Analysis of the Corpus
1.4.2.1 Approaches
Corpus analysis was the mainly used approach in this thesis to investigate
discourse features of hedging expressions. A corpus (sample) of naturalistic
economic discourse excerpts that are relevant to the thesis’s particular research
questions was collected. The discourse corpus was analyzed by counting the
frequency of discourse elements, categories, sequences, and combinations of
these linguistic/ discourse entities. The frequencies could be normalized by
counting the number of occurrences per 100 words. According to the hard-core
scientific researchers, the discourse corpus needs to be sampled systematically
and scientifically, rather than haphazardly or with bias. To avoid possible
mistakes, the corpus for the thesis was selected carefully, at random and with
suggestions of economic professors.
1.4.2.2 Methods
A comprehensive understanding of hedging in English and Vietnamese ERAs
involves at least two levels of linguistic description: an analysis at surface-level
hedging devices employed in a particular academic genre, and a deeper
functional analysis at pragmatic level. The analysis of the corpus was undertaken
by blending qualitative and quantitative methods. The analysis followed several
steps, beginning with a quantitative search, followed by a computerized
compilation of the corpus, then by a qualitative analysis all the texts and a

1.4.2.3.1 Compiling a list of hedging devices.
Firstly, a list of frequent items used to express mitigation was compiled by
reference to published studies about epistemic modality, verbs, adverbs…
introduced by Holmes (1988), and Kennedy (1987). For example, Holmes’s work
indicates the likelihood of encountering epistemic verbs such as appear, assume,
suggest, or suppose in academic discourse. Kennedy shows that tentative
expressions such as seldom, infrequently, or hardly ever In the most distinctive
or significant features of hedging devices, lexico-grammatical features and
structure interpretation were mostly focused on. Hedging devices and their
contexts for later reference were stored in computer, with codes indicating the
linguistic status of the items, the title of the article it belongs to, its position in the
text and other contextual categories. This phrase of the analysis was the surface
level analysis of hedging phenomena.
To avoid the trouble of quoting a whole text every time an example is needed, all
50 texts were divided into numbered sentences and each sentence was divided
into clauses (see Appendix 2 for more detail). Moreover, because of this
dividing, the linguistic context from which a particular quotation was extracted in
the appropriate text would be found easily. Another, and probably more
important reason is that when sentences and clauses are divided into minimal
linguistic unit, some semantic relations of a proposition or a set of propositions
will probably seen clearly.
- 16 -

Based on the verb or predicate, clauses were identified. Every clause was
assumed to have a verb which is either finite or non-finite (with –ing form verb, -
ed form verb or to-infinitive verb). Regarding to verbs which were elliptically
omitted, they were considered to be restored in the clause without affecting the
meaning of the sentence including that clause. For example,
(2a) In May, the three months’ interbank rate stood near 10% (2b) and
the long-term government bond rate at 12% (EE. 14)