Open Journal of Modern Linguistics
2013. Vol.3, No.2, 119-126
Published Online June 2013 in SciRes (http://www.scirp.org/journal/ojml) http://dx.doi.org/10.4236/ojml.2013.32016
Evaluative Functions of Reporting Evidentials in English
Research Articles of Applied Linguistics
Linxiu Yan g
Foreign Languages School, Shanxi University, Taiyuan, China
Received September 20th, 2012; revised February 27th, 2013; accepted March 5 th, 2013
Copyright © 2013 Linxiu Yang. This is an open access article distributed under the Creative Commons Attribu-
tion License, which permits unrestricted use, d i s tribution, and reproduction in any me di um, provide d the original
work is properly cited.
Reporting evidentials are frequently used in Research Articles. Based on the data analysis of 50 English
research articles of applied linguistics, the study shows that reporting evidentials not only function as in-
dicating the information sources, but also have multiple evaluative functions. The analyses have proved
this by showing the evaluative functions of reporting evidential in choosing different information sources
and different realization forms. At the same time the persuasive effects and discourse implications of
these different choices are also discussed.
Keywords: Reporting Evidential; English Research Articles; Evaluative Functions
Introduction
As a hot research issue in recent years, evidentiality has been
studied from various perspectives (e.g. Aikhenvald, 2003, 2004;
Chafe, 1986; Palmer, 1990, 2001; Mushin, 2000, 2001; Halli-
day & Matthiessen, 2004; Hu, 1994, 1995; Fang, 2005; Tang,
2007; Yang, 2009, 2010). With different research focuses, goals
and perspectives, these studies have provided us different un-
derstandings of evidentiality. Yet, up to now, few researchers
have touched the evaluative functions of evidentiality, espe-
cially the functions of reporting evidential. As the frequently
used evidential type in English research articles, it is necessary
to study what reporting evidentials can do for the writer. There-
fore, to fill this gap, this study intends to focus on the evalua-
tive functions of reporting evidentials in English research arti-
cles, aiming to show how reporting evidential can help the wri-
ters to negotiate the relationship among the information, the
writer and the reader.
Understanding of Evidentiality in the
Current Study
As for what is evidentiality, there has been no consensus.
There are narrow and broad understandings of it. For the work-
ing definition of evidentiality, the current study adopts the
broad view of evidentiality.
First, it treats evidentiality as a semantic notion rather than
a grammatical one and admits all the potential realization
forms rather than the grammaticalised ones. It adopts the
“one-to-many” approach in Systemic Funtional Linguistics and
admits the differences in different realizations for the same
semantics.
Second, the study agrees that evidentiality is interpersonal by
nature and negotiating the interpersonal relationship is one of
the most important functions of evidentiality, but at the same
time it holds that the interpersonal functions of evidentiality are
context-dependent. Only in certain concrete context, can the
interpersonal functions of a certain evidential be decided. For
instance, the reporting evidential it is said may perform differ-
ent interpersonal functions in different contexts. It may denote
the speaker’s uncertainty of the source of saying, or it is a de-
vice for the speaker to conceal the information source and dis-
tance him or her from the information, or even escape from
taking responsibility from the information. In this sense, con-
text is much important in deciding the interpersonal functions
of evidentials.
Third, evidentiality is much related to genre convention.
Many factors will affect the adoption of evidentiality, and genre
is undoubtedly one of them. Each genre has its own linguistic
manifestations. As far as evidentiality is concerned, in different
genres, the forms and distributions of evidentials are different.
For example, such subjective evidentials as I think, in my opin-
ion are not preferred in academic genres.
These understandings concerning evidentiality will decide
the analytical orientation in the later part. This is also a starting
point for the current research. It will examine in RAs what in-
terpersonal functions reporting evidentials will perform and
how they can help the writers to negotiate the relationship a mong
the information, the writer and the reader.
Data and Methodology
English RAs of applied linguistics are chosen as the data.
The corpus consists of 50 RAs in applied linguistics amounting
to about 350,000 words. The journals selected for this study are:
Journal of English for Academic Purposes (2004-2008), Jour-
nal of English for Specific Purposes (2004-2008), and Journal
of Pragmatics (2004-2008). The data of RAs are confined to
the same period because of the fact that genres are on the one
hand quite stable in a certain period of time. On the other hand,

L. X. YANG
they are also in a state of constant evolution, as Fairclough
(1992) notes, “a genre implies not only a particular text type,
but also particular processes of producing, distributing and con-
suming text… Changes in social practice are both manifested
on the plane of language in changes in the system of genre, and
in part brought about by such changes”. The genre of RAs also
may change over time. Therefore, in order to examine the lin-
guistic features of RAs, the study chooses RAs published dur-
ing the same time for the validity of the research results.
The data-coding of this research is done manually at the pre-
liminary stage to identify and count all the potential lexical and
discourse-based items that indicate different reporting eviden-
tial type s. The mate rial for data-coding includes the body of the
articles, i.e. the complete text of the articles, excluding abstracts,
notes, linguistic examples, tables, and figures. Then, Microsoft
Office Excel is adopted to deal with the data and draw the fig-
ures accordingly. In addition, in order to take the context of
evidentials into consideration to find the concordance patterns,
a concordance software is also adopted. This quantitative ap-
proach is meant to identify the frequency of occurrences and to
produce comparable data. The frequency of occurrence of each
group of items is calculated in permillage.
Classification of Reporting Evidential and Its
Lexicogrammatical Realizations in English RAs
Based on the genre convention of English RAs, in the current
research, reporting evidentials are classified into two types ac-
cording to the information source: other-reporting and self-re-
porting evidentials. Self-reporting evidentials indicate that in-
formation comes from whatever related to the writer’s own
research, e.g. I, we, our, my, our analysis, our research, this
article, and the participants involved in the experiments and so
on, while other-reporting evidentials indicate that information is
from the extra sources other than the writer’s own research.
Our data survey indicates that reporting evidentials are the
most important and frequently adopted evidentials. They have
various types of realizations.
First, (author + date) form is a conventional way to realize
reporting evidentials. For example:
1) Such evaluations can be said to be averrals which are
expressed as though deriving from a source, in this case,
implied consensus (Hunston, 2000).
In Example (1), the evidential (Hunston, 2000) indicate the
information comes from Hunston. At the same time it provides
a way for the writer to give a summary or generalization of the
cited information. This type of evidentials is typical in RAs.
This type also includes (website) which indicates the infor-
mation source is a certain website rather than an author. For
example:
2) Negative judgments can also be made implicitly, with
absence of items that carry negative values, but with to-
kens that evoke negative judgements from readers (http://
www.rammatics.com/appraisal/appraisalGuide/UnFrame
d/stage2-Attitude-judgment.htm).
Example (2) reveals that the internet provides an alternative
source of information. However, this type is not very frequently
adopted in the data and only several cases are found.
Second, reporting evidentials can be realized by verbal forms:
verb + that structure, be verbed structure, it is verbed structure
and as structure.
The structure verb + that is a way in which the writer can
show the specific information source, either human or non-
human, specific or unspecific. This form presents the informa-
tion source as the theme, which foregrounds the information
source rather than the cited information. For example:
3) Tannen has demonstrated that controlling others in-
volves them in a relationship (power entailing solidarity),
the same way that claiming intimacy has an element of
control (solidarity entailing power).
4) This body of literature suggests that L2 learners’ rela-
tionship with their advisors dramatically impact their par-
ticipation in academic literacy projects and, by extension,
their attempts to gain admittance into target discourse com-
munities.
5) Belcher’s research suggests that a critical factor in
high-level academic literacy activities is the quality and
kind of relationship that L2 learners develop with their
advisors.
6) Many researchers have argued that genre knowledge
plays a pivotal role in advanced academic litera cy.
In the above examples, by foregrounding the information
sources, the writers put more value on the information sources
rather than the information itself, which shows the writers’
respect for other researchers. The examples also show that the
information may be human, as in (3) and (6), or inhuman, as in
(4) and (5). It may be specific, as in (3) and (5), or unspecific,
as in (4) and (6).
The structures of be verbed and it is verbed allow the writer
to omit the information source for whatever reasons. Consider
the following two examples.
7) It is assumed that established genres such as case his-
tories, experimental research reports and editorials con-
stitute a natural part of readings in the medical sciences.
8) As can be seen in examples below, the DM te was
found to function mainly as an information state marker
and mostly marked shared and assumed knowledge be-
tween the speaker and the addressee.
In Examples (7) and (8), instead of explicitly indicating the
information sources, the writers choose to conceal them. In this
case, the writers pay more attention to the reported information
rather than where the information comes. The writers may not
know the information source or they find no necessity to point
it out. What they value is just the cited information, which is
different from the case of the structure of verb+ that.
Reporting evidentials can also be realized by as verb(ed)
structure. Some examples of as structure are given below.
9) However, as Hyland (1998a) adds, expressions of cer-
tainty work towards the acceptance of by addressing
readers as knowledgeable peers who are familiar with the
ideas presented and able to follow the author’s reasoning.
10) Moreover, as noted above, these labels are interactive:
their use affects the reader’s perception of the proposi-
tions and so enables the readers to perceive the organiza-
tion and meaning that the writer intends.
As verb(ed) structure is often chosen by the writer because of
the flexibility of its occurrence. It can occur either at the begin-
ning, in the middle or at the end of a clause. It is also a kind of
120

L. X. YANG
textual meta-discourse which can smooth and guide the reader’s
understanding of the writer ’s argumentation, as in Example (10).
This point will be elaborated in the following chapter.
Third, non-verbal reporting evidentials include noun pat-
terns or adjuncts. The typical nouns and adjuncts are such as
fact, observation, agreement, finding, view, claim, evidence, ar-
gument, suggestion, according to X, in X’s data, in X’s view, in
X’s terms and so on.
In this type of realization, the most frequently-used noun is
fact which nearly constitutes 90% of the nouns as reporting
evidentials, as shown in Example (11).
11) The use of nouns to construct stance in academic
writing has so far attracted little attention, despite the fact
that several researchers have identified a group of nouns
which offer the possibility of incorporating interpersonal
meanings in the text.
The use of nouns as reporting evidentials have its own pecu-
liar functions and characteristics. In some cases, the informa-
tion source may be concealed. In addition, this form can also
provide the writer with chances to evaluate the information
source, e.g. One further interesting finding, the most striking
finding for Harwood, Thomson’s groundbreaking study. Fur-
thermore, nouns allow the process to be a participant which can
not be argued, negated and so on. Therefore, they have more
persuasive power and they make it more possible for the reader
to accept what the write r expre sses.
The adjunct according to is frequently adopted to indicate
information source, which typically occurs at the beginning and
gives prominence to information source. This type of realiza-
tion is very objective because it just indicates information
source without any of the writer’s evaluation of information
source and cited information.
12) According to Swales (1996), these are genres that
“operate to support and validate the manufacture of
knowledge directly as part of the publishing process itself
or indirectly by underpinning the academic administrative
processes of hiring, promotion and departmental review.”
In Example (12), the writer chooses the adjunct “according
to” to indicate the information source and shows no evaluations
of the information source and the information itself. This type
of reporting evidentials are identical to the objective nature of
RAs.
Table 1 will present a clear picture of the lexicogrammtical
realizations of reporting evidential in English RAs of Applied
Linguistics.
Evaluative Functions of Reporting Evidential
Based on the descriptive result of Section 3, in this section,
we will look at the evaluative functions of reporting evidential
in four aspects: the phraseological patterns of reporting eviden-
tial, the evaluative functions of information sources, the evalua-
tive functions of reporting verbs and evaluative functions of
nouns.
Phraseological Patterns of Reporting Evidentials
First consider the distribution pattern of reporting evidentials
in RAs, as is shown in the Table 2.
As seen from Table 2, to express reporting evidentiality, the
writer prefers verbal forms, either in active forms or in passive
forms. Verbal forms nearly constitute 61.4% of all the realiza-
tion forms. The second frequently used forms are (author +
Table 1.
Lexicogrammatical realizations of reporting evidential s.
Evidential type Realization type Lexicogrammatical realizations Typical examples
(Author + year) or (website + year) (Hunston, 2000)
Verb that structure, be verbed structu re, X argue, maintai n , fo und, … that
It is ved structure It is argued, it has been revealed Verbal realization
As structure As indicated by…
Noun that Fact, observation, agreement, finding, view, claim,
Reporting
evidential ty pes
Non-verbal realization Adjunct According to X, in X’s data, in X’s view
Table 2.
Distribution of reporting evidentials.
Realization forms Other-reporting Self-reporting Total & percentage
Author/date 647 0 647 30.7%
Verb that structure 400 550 950
(It) is ved structure 53 50 103
Verbal
As structure. 70 107 177
61.4%
Noun that 117 3 120
Non-verbal Adjunct 37 10 47 7.9%

L. X. YANG
date) forms, constituting 30.7%. Therefore, verbal forms are
prominent in reporting evidentials. The effect of this fore-
grounding feature will be discussed in Section 4.3. The results
also show that (author + date) convention is a very specific
form of reporting evidentials in RAs. It generally occurs at the
end of a proposition with the only purpose of indicating the
information source of that proposition. For example:
(13) It sends the message to teachers that voice is criti-
cally important, and this message, if passed down to stu-
dents, may result in learners who are more concerned with
identity than ideas (Stapleton, 2002).
In Example (13), to show that the proposition “It sends the
message to teachers that voice is …” comes from the other
source other than the writer himself, he chooses (author + date)
form. Compared with other forms, (author + date) forms are
the objective ways to present the information sources in that the
writer only reveals where the information is from without any
subjective intrusion into the proposition. The writer will leave
his “imprint” on choosing how to indicate the information
sources. When he chooses (author + date) forms, he will be
quite distanced from the proposition. It is the cited authors, but
not the writer, who bear the full responsibility for the validity of
the proposition. This kind of form is also identical to the objec-
tive nature of RAs in which not too much subjectivity is in-
volved.
Another important point peculiar to (author + date) form is
that by indicating the exact date, sometimes with exact page
number (e.g. Stapleton, 2002: p. 187), the reliability of the
propositions is greatly improved. It can be certain enough that
the reader tends to believe the information with specific sources.
Therefore, this form contributes much to the persuasive and
rhetorical purposes of the whole genre. First, it can improve the
reliability of the information. Second, the improved reliability
adds to the credibility of the writer. The writer will be made
more detached from the information presented, therefore with
less commitment and responsibility for the validity of the in-
formation. Third, the form is almost the most objective way for
the writer to present the information from other sources, which
consolidates the objectivity of RAs.
As shown above, (author + date) forms are an objective way
to function as reporting evidentials, but it is a different picture
for verbal and noun forms as evidentials. The choice of report-
ing verbs and nouns will show the writer’s subjective evalua-
tion of the reported information and also the information
sources. Sections 4.3 and 4.4 will elaborate the eval uative func-
tions of reporting verbs and nouns as reporting evidentials.
Evaluative Functions of Information Source
In addition to the evaluative potential in different phraselo-
gical patterns of reporting evidentials, the choice of information
source is also evaluative in function and related to the overall
persuation of RAs.
As can be seen in Ta b le 3 , specific human sources are most
frequently chosen as information sources. Non-human and con-
cealed sources are relatively low in frequencies. However, the
different distributions between other-reporting and self-report-
ing should be devoted much attention to. In other-reporting, the
writer tends to choose specific human sources which constitute
88.6% of the total number. This shows that in presenting others’
work, the writer gives much prominence to the cited authors
than the cited information. In so doing, there are at least two
persuasive effects. First, by giving prominence to the cited au-
thors, the writer will show his respect for the previous related
researchers, which helps to build a professional persona. Sec-
ond, this strategy adds much to the reliability of information
and also the credibility of the writer. Nesler et al. (1993) points
out that people tend to accept beliefs, knowledge, and opinions
from what they see as authoritative, trustworthy, or credible
sources, such as scholars, professionals, experts or reliable me-
dia. In addition, with reference to specific sources, the reliabil-
ity of information will be improved. For example, Hu (1994)
points out that a specific source will add to the reliability of
information because the reader has specific persons and sources
to refer to.
The situation is different for self-reporting evidentials. The
prominent information source adopted by the writer is non-
human source (62.7%), including the findings, data, figures,
analysis, tables and so on. This is different from other-reporting
evidentials. However, they have the similar ultimate purposes.
When presenting his own work, in order to let the facts speak
for themselves, the writer tends to choose the research as the
information sources, such as the research shows rather than we
show. In so doing, the reliability of information is improved
and at the same time contributes to the objectivity of RAs be-
cause not so many Is and wes are involved. When the writer
chooses human sources as information sources, in addition to Is
and wes, it is worth noting that the participants in the research
are chosen as the information sources, e.g. the interviewees and
the research participants. This also adds to the reliability of the
information, for these people are direct experiencers and eye-
witnesses of the information presented.
The above discussion has shown that choosing different in-
formation sources is also meaningful and evaluative. It also
presents the differences between other-reporting evidentials and
Table 3.
Information s ou rces of reporting evidentials.
Information sources Other-reporting Self-reporting Total report ing
Specific 1037 88.6% 83 14.3% 1120 64%
Human Unspecific 7 0.6% 13 2.3% 20 1.1%
Non-human 37 3.2% 363 62.7% 400 22.9%
Concealed 90 7.6% 120 20.7% 210 12%
Total 1171 100% 579 100% 1750 100%
122

L. X. YANG
self-reporting evidentials in this respect. In spite of the differ-
ences, they both serve the ultimate persuasive purpose of RAs.
Evaluative Functions of Verbs as Reporting
Evidentials
Verbal forms are the most frequently adopted in RAs to func-
tion as reporting evidentials. By adopting verbal forms, the
writer does not just show where the information is from. In-
stead, he presents his subjective evaluation. This section will
focus on the evaluative functions of reporting verbs.
In RAs, reporting verbs do not simply function to indicate
the sources of the information reported, but they also reveal the
writer’s own position. The selection of an appropriate reporting
verb allows the writer to intrude into the discourse to signal his
assessment of the evidential status of the reported proposition
and to demonstrate his commitment. For example, the verbs say
and insist in He says/insists he is innocent differ in discourse
implications in that insist explicitly conveys the speaker’s in-
sistence on the part of the information presented.
Sometimes, reporting verbs are used with adverbials (e.g. As
sb correctly asserts). This explicitly evaluative strategy allows
the writer to open a discursive space within which the writer
either exploits his opposition to the reported message or to
build on it. However, this is not a common case in RAs. Most
of the time, the writer chooses to intrude into the proposition
implicitly.
Hyland (1999: pp. 349-350) finds that reporting verbs in aca-
demic texts, such as X observe, X advocate, X establish, X ig-
nore, X fail, and so on, can help the writer to differentiate his
various degrees of commitment to the cited messages, which at
the same time demonstrate implicitly the writer’s personal stances
towards the cited authors.
Thompson (1991) agrees with Hyland on the evaluative
functions of reporting verbs in academic discourse. As Thomp-
son & Ye’s (1991) ground-breaking study shows, the choice of
reporting verbs is a key feature which enables the writer to
position his work in relation to that of other members of the
discipline. Their study offers a threefold analysis of the evalua-
tive potential of reporting verbs (Thompson & Ye, 1991: pp.
372-373). First, reporting verbs show the cited author’s stances
towards the information. Second, reporting verbs can construct
the writer’s stance of acceptance, neutrality or rejection of the
cited research. Third, they allow the writer’s interpretation of
the author’s behaviors or discourse.
Thompson and Ye (1991) also distinguish three categories of
reporting verbs according to the activities they perform: “tex-
tual” verbs in which there is an obligatory element of verbal ex-
pression (e.g., state, write); “mental” verbs, which refer to men-
tal processes (e.g. think, believe); and “research” verbs, which
refer to the processes that are part of research activity (e.g. find,
demonstrate).
Later studies such as those by Thomas & Hawes (1994) and
Hyland (2002) also admit the evaluative functions of reporting
verbs. Hyland’s category of reporting verbs, which are diverged
from Thomson and Ye’s rather complex system, is shown in
Figure 1. The figure clearly shows Hyland’s opinion toward
the evaluative functions of reporting verbs.
The above has shown that many researchers have paid much
attention to the evaluative functions of reporting verbs. The use
of a reporting verb to introduce the work of others is also a
significant rhetorical choice (Hunston, 1993; Tadros, 1993;
Figure 1.
Categories of report i ng v e rbs (after Hyland, 1 99 9: p . 350).
Thomas & Hawes, 1994; Thompson & Ye, 1991). The impor-
tance of these verbs lies in the fact that they allow the writer to
convey clearly the kind of activity reported and to distinguish
an attitude to that information, signaling whether the claims are
to be accepted or not.
This study totally agrees the evaluative function of reporting
verbs. The following will discuss the evaluative functions of re-
porting verbs as evidentials. To categorize reporting verbs, the
book adopts the classification of Francis et al. (1996) for V that
clause pattern. According to his classification, three groups of
reporting verbs are categorized in our corpus: ARGUE group,
THINK group, SHOW and FIND group. The explanations and
verb samples are adapted from Francis et al. (1996: pp. 97-101),
as shown in the following:
A: ARGUE verbs are concerned with writing and other
forms of communication, e.g., argue, suggest, point out,
write, conclude, claim, add, maintain, propose, imply,
mention.
B: THINK verbs are concerned with thinking, including
having a belief; knowing, understanding, hoping, fearing,
e.g., think, assume, feel, hold, believe.
C: SHOW and FIND verbs are concerned with indicating
a fact or situation or with coming to know or think some-
thing, e.g., show, demonstrate, reveal, find, observe, dis-
cover, indicate.
It is important to note that the categorization of verbs is de-
pendent on the context where the verbs occur. That is, a verb
can occur in more than one group, and the context needs to be
examined in order to determine the appropriate category the
verb belongs to. For example, the verb observe can appear in
FIND group when it refers to the visual evidence with the
meaning of “noticing” and also can be in ARGUE group when
it refers to the language activit y.
This categorization may overlap or be similar to those of

L. X. YANG
Thompson and Ye’s and Hyland’s. ARGUE verbs parallel the
textual group and discourse group, THINK group, the mental
and cognition group and SHOW and FIND group, the research
group. In spite of the similarities and overlaps, the categoriza-
tion adapted from Francis et al. (1996) is chosen by the current
study for the following two reasons. First, this categorization
reveals better the ways information is acquired, either through
language activity (ARGUE verbs), through vi sual channel (FIND
verbs), or through thinking (THINK verbs). Second, verbs of
different categorizations may denote a line of different com-
mitment and certainty. For example, FIND and SHOW group
tend to bear high certainty than the other two groups in that
FIND and SHOW verbs are always factive. They indicate dif-
ferent degrees of reliabili ty of information.
Based on the above categorization, a statistic picture of re-
porting verbs in NS corpus is presented in Table 4.
Table 4 shows the distribution of verb groups in reporting
evidential type. There are great differences in the distribution of
verbal groups and also between other-reporting evidential type
and self-reporting evidential type. It shows that THINK verbs
are seldom adopted as reporting evidentials. Especially in self-
reporting evidential type, no THINK groups occur, which can
be explained by the fact that this book categorizes THINK group
with self sources as brief evidentials. However, for the other
two types of verbs, significant differences are presented. In
other-reporting evidential type, ARGUE verbs predominate. In
fact, 61.1% of the total appear with an ARGUE verb. The most
frequently adopted ARGUE verbs are argue, point out, suggest,
claim which constitute nearly half of the total number, as shown
in the following examples.
14) Swales (1990) suggests that citation convention (nu-
merical or author/date) may affect the choice between in-
tegral and non-integral and he argues that numerical
conventions predispose the writer to use non-integral cita-
tion.
15) As Johns and Swales (2002: p. 13) point out, uncer-
tainty over “what role there might be for a personal voice”
is one area of difficulty that affects student writing at all
levels, including the thesis.
The situation is quite different for self-reporting evidentials.
Unlike other-reporting evidentials, FIND and SHOW verbs
predominate with the occurrence of 73.6% of the total. FIND
and SHOW verbs are mainly concerned with the writer’s own
researches, such as the results, situations, findings and analyses.
For example:
16) Our study of conversation in noisy settings shows
that there are also identifiable patterns in the ways that
noise and impaired language perception during conversa-
tion affect grammatical and discourse structures, language
processing, language use, and patterns of interaction in
conversations.
17) Our data has demonstrated that the acoustic con-
straints have clear repercussions on grammatical con-
structions, including effects on utterance lengths, gram-
matical complexity, and questioning strategies.
The two examples above show that the information sources
of self-reporting evidentials are mainly about the writer’s own
study such as our data, our study and so on. In fact, our study
has found very low percentage of personal pronouns such as I
and we for information sources. This finding is also different
from that of other-reporting evidentials, which can be explained
by the fact that when referring to other sources, the writer tends
to give prominence to the cited authors themselves, while when
referring to self sources, he will put more value on the studies
rather than the writer himself. This is a persuasive strategy.
When presenting his own studies and researches, the writer lets
his studies and researches speak for themselves, but not his own
subjective demonstration. This strategy of “objectiveness” adds
to the reliability of information. Thus, the reader will be more
likely to accept the claims the writer makes.
To sum up, the choice of reporting verbs in reporting eviden-
tial type positions the writer in relation to the reported authors
and the reported information. With different reporting verbs, the
writer shows his evaluation and stances towards the reported
authors and the reported information.
Evaluative Functions of Nouns as Reporting
Evidentials
The above has shown that (author + date) and reporting
verbs are often adopted in reporting evidential type in RAs. In
spite of this, significant numbers of nouns as reporting eviden-
tials (e.g. fact, finding, evidence, suggestion, and observation)
also occur in the data. These nouns often occur with that-clause
Table 4.
Distribution of the verb groups in reporting evidentials.
Other-reporting Self-reporting
Verb groups % of total The most frequent
verbs (total number of occurrence) % of total The most frequent verbs (total number of o ccurrence)
ARGUE grou p 61.1
Argue 70
Point out 43
Suggest 40
Claim 37
26.4
Suggest 70
Explain 33
Note 23
State 20
THINK group 3.8 Assume 10 Think 3
Hold 3 0 0
FIND and SHO W group 35.1
Show 43
Note 20
Reveal 17
Observe 13
Indicate 10
73.6
Show 117
Find 100
Reveal 63
Indicate 53
Demonstrate 13
124

L. X. YANG
(e.g. Swales’ suggestion that review of literature does not only
occur…). It is thought that the use of noun with that-clause as
reporting evidentials is worthy of study for their distinct eva-
luative functions. This pattern enables the writer to give his
evaluations of the propositions following that. Nouns as evi-
dentials have their specific advantages. In fact, Biber et al.
(1999) have stated that nouns are one of the primary devices
used to express the writer’s evaluation and stance in academic
writing. Based on the close examination of this pattern used in
the corpus, their evaluative functions will be shown in the fol-
lowing, which will indicate how theses nouns can help the
writer express his own evaluations and construct a convincing
argument.
First, evaluative functions of nouns as evidentials are real-
ized by choosing appropriate nouns. The use of Noun that-
clause pattern encapsulates the proposition in the clause, sum-
marizing and representing it to the reader, which enables the
writer to incorporate his own evaluations of the propositions
through the choice of different types of nouns. Biber et al.
(1999) find that the evaluative functions of nouns may either be
attitudinal or epistemic. They show the writer’s evaluation of
and different degrees of commitment to the information. For
example, the use of “suggestion” is very different from the
nouns of “claim”, “findings”, or “fact”.
Thus, in order to understand the evaluative functions of
nouns as reporting evidentials, it is necessary to examine the
nouns that occur in this pattern. To categorize these nouns, the
dissertation adopts the semantic criteria by Francis et al.
The categories of nouns as evidentials are as follows:
Fact and Findings Group: these nouns refer to the facts, or
the findings in the research (e.g. fact, findings, observation).
Idea group: these nouns refer to belief, ideas, wishes, and
thought processes (e.g. suggestion, idea, assumption, view, be-
lief).
Argument Group: these nouns refer to something that is
written or spoken (e.g. argument, claim, point, agreement).
Table 3 shows the different frequencies of the occurrence of
every category of nouns as evidentials in the corpus.
Table 5 shows that there are significant differences among
the categories of nouns as reporting evidentials. The writer par-
ticularly favors the noun category of Fact and Finding which
nearly constitutes 73% of the total number of nouns as eviden-
tials. This result shows that by using nouns indicating Fact and
Finding, the writer puts more value on the factual status of the
information presented. If something is presented as a fact or a
finding, a reader is more likely to accept it. Therefore, this type
of knowledge may occupy higher reliability than the other two
types. In this sense, choosing an appropriate noun is critical for
the reader’s acceptance of the claims.
Second, nouns as reporting evidentials provide the writer
with a relatively objective way to evaluate. Different from other
Table 5.
Categories of nouns as evidentials.
Noun category Raw data Frequency per 1000 words
Fact and finding 87 0.277
Idea 23 0.075
Argument 10 0.032
Total 120 0.384
types of reporting evidentials, nouns as evidentials can facilitate
the construction of a seemingly ‘objective’ evaluation of the
proposition in the that clause since the writer can avoid indi-
cating the source of the proposition just as verbs as evidentials
do (excluding the passive forms of verbs). In fact, the data has
shown that the majority of the nouns as evidentials do not occur
with specific names or personal pronouns for the purpose of
attribution. Actually only 6 cases of nouns are found to occur
with the sources, which comprise 16.7% of all the cases. The
absence of information sources enables the writer to obscure
the origin of any evaluation that is carried out. Therefore, the
information appears objective and is less open to dispute.
Third, evaluation of nouns as reporting evidentials is multi-
layering. Nouns as reporting evidentials are sometimes without
a head. For example, with no specific names or personal pro-
nouns to indicate the information sources, they have the func-
tion of multi-layering evaluation. By multi-layering, it is ex-
plained by reference to the notions of attribution and averral
(Sinclair, 1986; Tadros, 1993). According to Sinclair, a text is
made up of propositions which may be put forward by the
writer (averrals) or attributed by the writer to some other person
or entity (attributions). In RAs, all assertions are taken to be
averrals, unless the writer clearly shows the source of the asser-
tions. In making an averral, the writer is responsible for the
veracity of the proposition advanced. Consider the following
examples.
18) Specifically, the use of parentheses by the Spanish
writers (p = 0.0016) put forward the idea that Spanish
opinion columns may exhibit a greater freedom to include
what the English-speaking rhetorical principles consider
“supplementary or digressive” material.
19) As far as the limitations observed by Samraj (2002)
are concerned, the introduction of an optional step (step 2),
“presenting positive justification”, in Move 2 accounts for
her first criticism, and Swales’ (2004) suggestion that re-
view of literature does not only occur throughout the in-
troduction but can occur throughout the article as a whole
accounts for her second criticism.
In Example (18), no information source is indicated. Thus, it
is the writer who is responsible for the truth of the statement
and it is the writer who holds the idea of “Spanish opinion
columns may exhibit a greater…”. However, in (19), the infor-
mation source is clearly indicated through “Swales’”. It is
Swales not the writer who takes the responsibility for the state-
ment “review of literature does not only occur throughout…”.
However, it is only superficially so. As Sinclair (1986) points
out, it is the writer who bears the ultimate responsibility for all
the propositions in his texts. In this sense, all the attributions
are also averrals. It is the writer who bears the responsibility for
the whole statement “As far as the limitations observed…” and
for the choice of the noun “suggestion”. In this case, superfi-
cially, the information source “Swales” seems to be responsible
for the statement. Actually the writer’s own evaluation is also
incorporated in it. For example, he chooses the noun “sugges-
tion” rather than “claim” or “finding”, which shows that even in
attribution, the writer’s evaluation is also revealed.
Thus, the evaluation of nouns as reporting evidential is
sometimes multi-laying. In such cases, writers show their
evaluations which they assign to others or entities, but which
simultaneously express their own positions. This multi-layering

L. X. YANG
of evaluation provides a resource which writers can adopt to
incorporate their own evaluations, while appearing to report
that of others.
Fourth, nouns as evidentials make a verbal process an entity
or phenomenon, such as “suggestion” rather than “suggest”. It
appears that what the writer presents is something that exists in
the world and it is more likely for the reader to accept what has
existed in the world than what others say.
Fifth, in addition to the above evaluative functions and
characteristics of nouns as reporting evidentials, it is also found
that in some examples the nouns are modified by attributives
such as “general finding, the most striking finding for Harwood,
another important, though not surprising finding” and so on.
Allowing different modifiers to modify nouns as evidentials
may also be a great advantage for the writer to add explicit
evaluations toward propositions.
In sum, nouns as evidentials have great evaluative potential
and provide the writer with an alternative to present the infor-
mation with his own evaluations and stances.
Conclusion
This study has shown that in construing reporting evidential
type, the writer has great power because he has various linguis-
tic forms to choose for his presentation. How the writer chooses
to present information is as important as the information he
wants to present. Just as Berkenkotter & Huckin (1995) say,
“you are what you cite”. No matter what kind of forms he
chooses, the writer leaves his imprint there and expresses his
evaluation or stance. The analyses have proved this by showing
the evaluative functions in choosing different information sources
and different realization forms. At the same time it shows the
persuasive effects and discourse implications of these different
choices.
This study intends to help to raise the writers’ awareness in
choosing reporting evidentials in RA writing. Theoretically, it
is a beginning to study what evidentiality can do for the lan-
guage users other than indicating the information source. It may
lay a foundation for the future research and provide orientation
for further study. There are more areas to be further studied.
First, the functions of other evidential types can be further
studied; Second, because of the genre convention, evidential
use in other genres, even evidential use across genres is worthy
of more research; Third, evidential use in different cultures may
vary, which is believed to be an interesting topic in evidential
study.
Acknowledgments
This work was supported by Chinese Educational Bureau
[grant number: 11YJC740128] and Program for the Outstand-
ing Innovative Teams of Higher Learning Institutions of Shanxi
(OIT).
REFERENCES
Aikhenvald, A. (2004). Evidentiality. Oxford: Oxford University.
Aikhenvald, A., & Dixon, R. (2003). Studies in evidentiality. Amster-
dam/Philadelphia: John Benja mins Publishing Company.
Berkenkotter, C., & Huckin, T. (1995). Genre knowledge in discipli-
nary communication. Hillsdale, NJ: Lawrence Erlbaum.
Biber, D., et al. (1999). Longman grammar of spoken and written Eng-
lish. Edinburgh: Pearson Education Ltd.
Chafe, W. (1986). Evidentiality in English conversation and academic
writing. In W. Chafe, & J. Nichols (Eds.), Evidentiality: The linguis-
tic coding of epistemology (pp. 26 1-272). Norwood, NJ: Ablex.
Chafe, W., & Nichols, J. (1986). Evidentiality: The linguistic coding of
epistemology. Norwood, NJ: Ablex.
Fairclough, N. (1992). Discourse and social change. Cambridge: Polity
Press.
Fang, H. M. (2005). A Systemic-functional Approach to evidentiality.
Unpublished Doctoral Dissertation, Shanghai: Fudan University.
Francis, G., Hunston, S., & Manning, E. (1996). Collins COBUILD
grammer patterns 1: Verbs. London: Harper Collins.
Halliday, M. A. K., & Matthiessen, C. M. I. (2004). An Introduction to
Functional Grammar. London: Arnold .
Hu, Z. L. (1994). Evidentiality in language. Foreign Languages Teach-
ing and Research, 1, 9-15.
Hu, Z. L. (1995). Evidentiality in Chinese and discourse analysis.
Journal of Hubei University, 2, 13-23.
Hunston, S. (1993). Evaluation and ideology in scientific writing. In M.
Ghadessy (Ed.), Register analysis: Theory and practice. London and
New York: Pinter Publishers.
Hyland, K. (1999). Academic attribution: Citation and the construction
of disciplinary knowledge. Applied Linguistics, 20, 341-367.
doi:10.1093/applin/20.3.341
Hyland, K. (2002). Activity and evaluation: Reporting practices in aca-
demic writing. In J. Flowerdew (Ed.), Academic discourse (pp. 115-
130). London: Longman.
Mushin, I. (2000). Evidentiality and deixis in narrative retelling. Jour-
nal of Pragmatics, 32, 927-957.
doi:10.1016/S0378-2166(99)00085-5
Mushin, I. (2001). Evidentiality and epistemological stance: Narrative
retelling. Amsterstam/Philadelphia: John Benjamins Publishing Com-
pany.
Nesler, M. S., et al. (1993). The effect of credibility on perceived power.
Journal of Applied Social Psychology, 17, 1407-1425.
doi:10.1111/j.1559-1816.1993.tb01040.x
Palmer, F. (1990). Modality and the English modals. London: Long-
man.
Palmer, F. (2001). Mood and modality. Cambridge: Cambridge Univer-
sity Press.
Sinclair, J. (1986). Fictional worlds. In M. Coulthard (Ed.), Talking
about text (pp. 43-47). Birmingham: University of Birmingham ELR.
Tadros, A. (1993). The Pragmatics of text averral and attribution in aca-
demic texts. In M. Hoey (Ed.), Data, description, discourse (pp. 99-
114). London: Harper Collins.
Tang, B. (2007). Systemic-functional approach to discourse features of
evidentiality in English news reports of epidemic situation update.
Unpublished Doctoral Dissertation, Shanghai: Fudan University.
Thomas, S., & Hawes, T. P. (1994). Reporting verbs in medical journal
articles. Englsih for Specif ic P u rp os e s, 2, 134-155.
Thompson, G., & Ye, Y. (1991). Evaluation in the reporting verbs used
in academic papers. Applied Linguistcs, 4, 365-382.
Yang, L. X. (2009). Evidentiality in English research articles. Unpub-
lished Doctoral Dissertation, Xiamen: Xiamen University.
Yang, L. X. (2010). Genre perspective on evidentiality. Proceedings of
36th ISF. Sydney: Macquarie University Press.
126