Journal of Modern Languages
Vol. 30, No.2 (2020)
https://doi.org/10.22452/jml.vol30no2.3
Corpus Stylistic Analysis of Malaysian Online Columnists
Siti Aeisha Joharry
aeisha@uitm.edu.my
Akademi Pengajian Bahasa, UiTM Shah Alam
Syamimi Turiman
syamimituriman@uitm.edu.my
Akademi Pengajian Bahasa, UiTM Shah Alam
Abstract
Online media creates various platforms by which people can view and make sense of the
world today. In this paper, two Malaysian columnists from two national English online
portals: The Star Online and News Straits Times were selected for a corpus-assisted
stylistics discourse analysis. Frequency lists were firstly compared between each columnist
to identify salient words that are used by each writer. Initial observation shows that a
number of words refer to law/policy [e.g. act, law(s)] and government/public (e.g.
constitution, parliament). From the comparing wordlists feature, stylistic comparisons are
further explored using Hyland’s (2005) interactional metadiscourse features. The use of
the first-person pronoun ‘I’ was also investigated, which McNair (2008) regards as typical
of commentary journalism. Although findings show that both columnists employ similar
metadiscoursal features, Syahredzan projects a more assertive stance (I have, know) as
opposed to John Teo who is more suggestive in style (I think, believe). Results thus present
columnists’ style of writing, which are significant for readers when deciding on a piece of
news and to be critically aware of how persuasiveness can be constructed in journalistic
discourse.
Keywords: Corpus linguistics, media discourse, online newspapers, corpus-assisted
discourse analysis, stylistics
53
Corpus Stylistic Analysis of Malaysian Online Columnists
1.
Introduction
Corpus linguistics is an area of study that is expanding and its present-day popularity is due to the
interest in the techniques used, as it deals with “some set of machine-readable texts which is
deemed an appropriate basis on which to study a specific set of research questions” (McEnery &
Hardy, 2012, p. 1). In other words, corpus linguistics involves the use of computers to search and
analyse a collection of real language (Vaughan & O’Keeffe, 2015). The number of language
disciplines that utilize corpus linguistics has grown rapidly, which includes media discourse on
various issues such as newsworthiness (Bednarek & Caple, 2014), representation of Islam (Baker,
Gabrielatos & McEnery, 2013), migrants (Taylor, 2014), and feminism in media (Jarwoska &
Krishnamurthy, 2012).
At present, there are 5.11 billion mobile users, making online media the preferred platform
to access news and features on current affairs, among others. Online media allows for new ways
for readers to consume news and make sense of the world today. With the ability to simultaneously
reach a mass and highly dispersed audience, online media has developed into a setter of agenda
and moulder of public opinion (Djerf-Pierre & Shehata, 2017). This explains the increase of
journalistic commentary, and how events are reported through the lens of columnists.
News has been a common area for research (e.g. Bird, 2011; O’Keeffe, 2011), and language
of newspapers is identified as one of the four major registers in the English language, along with
spoken conversation, academic writing and fiction (Biber, Johansson, Leech & Finegan, 1999).
Since media discourse is public, it garners the attention of discourse analysts and increasingly so
because of the online availability of newspapers (O’Keeffe, 2011). While it is easy to obtain news
stories presented in the online newspapers, people reading the news pick and choose stories they
want to attend to and believe, and select from a seemingly endless supply of information to
assemble their own versions’ (Bird, 2011: p. 504). This highlights the role of columnists in the
information chain. Online columns have a significant role in reporting the news, presenting
informed summaries and interpretations of events and issues from their point of view to reach the
public. As McNair (2008) puts it, this type of journalism (also known as commentary journalism)
revolves around “making sense” of events and therefore their roles are more than mere reporting,
“but also a sifter and sorter of an accelerating, increasingly voluminous flow of information into
news items of differential value … [j]ournalism expanded its remit from the production of facts
… to the construction of meanings” (p. 106).
54
Siti Aeisha Joharry & Syamimi Turiman
However, in media discourse, the language used by columnists has yet to be examined
further –most linguistic studies on media discourse have investigated news/weblogs (e.g. Reese et
al., 2007; Baker et al., 2013; Bednarek & Caple, 2014). Studies on columnists’ writing in column
pieces tend to be underpinned by media study background (McNair, 2008; Meikle, 2008) whereas
linguistic analyses on opinion pieces are rather limited (except for a few: Fu & Hyland, 2014;
Noorian & Biria, 2010). According to Fu and Hyland (2014), contrary to typical news reporting,
opinion pieces take a more personal interactional position, adopting a clear perspective towards
both their topics and their readers by establishing a stance early on in the piece and supporting this
with a range of warrants for their opinions. What is key to opinion texts is the writer’s explicitly
subjective attitude and open judgments of the issues (Fu & Hyland, 2014, p. 4), which are therefore
interesting to be studied. To the best of the authors’ knowledge, a study has yet to explore how
Malaysian columnists construct their material and present this to readers, drawing on the resources
of interactional metadiscourse.
This leads to the present study for understanding the ways writers manage interaction by
intruding and commenting on their message –how they express solidarity, anticipate objections
and respond to an imagined dialogue with others (Fu & Hyland, 2014, p. 6). In today’s globalised
world, news is now easily obtained via the click of the mouse, particularly with the advent of the
internet. This brings about important value for investigating how interactions about news within a
text differ when they are transmitted via the online medium, and how they shape media discourse
of a particular society. As Hall (2008, p. 222) notes, “[j]ournalism’s dual role in the information
age is to act as an informed moderator for that interaction and the cartographer of the information
that feeds it”. Moreover, the integration of both discourse analysis and corpus linguistics
techniques has “identified conventionalised discursive devices that are repeatedly used in news
discourse to construct and perpetuate an ideology of newsworthiness” (Bednarek & Caple, 2014,
p. 40) that in turn, would render a piece of news reporting to be more persuasive and garner wider
readability.
So far, most research findings have resulted from non-linguistic studies of Western media.
This is argued to be not “equally applicable” (Bernstein & Galily, 2008, p. 181) to other countries
where dominant discourses that might be found in Asian media, such as Malaysia, might differ
from those of Western media due to distinct socio-cultural values and practices. As a result, the
current study is necessary to be carried out in Malaysia. This study thus aims to investigate
55
Corpus Stylistic Analysis of Malaysian Online Columnists
differences in writing style, particularly of two online columnists from two mainstream Malaysian
local online news portals, guided by the following research questions:
1.
What are the stylistic differences between two local journalistic commentary
writing?
2.
How does a corpus-assisted stylistic approach to media discourse able to highlight
salient patterns of writers’ rhetoric/writing strategies (metadiscoursal features)?
2.
Literature Review
The literature on corpus work in media discourse is extensive as more studies have investigated
language use in media via empirical methods like those offered in corpus linguistics (e.g. Bednarek
& Caple, 2014; Marchi, 2018). As a form of delineation, this section is divided into two: describing
past relevant studies in two corpus approaches –CADS or corpus-assisted discourse analysis, and
corpus-assisted stylistics.
2.1
CADS and Media Discourse
Apart from the traditional Conversation Analysis (CA), media discourse is also generally
examined using approaches such as Discourse Analysis (DA) and Critical Discourse Analysis
(CDA) (Fairclough, 1995; Wodak, 2001; & van Djik, 1993). These approaches are mainly
qualitative in nature, and are based on close readings of texts (Jarwoska, 2016). Both DA and CDA
are effective methods in the identification of norms and ideologies in media discourse through the
interplay between context, text and language in smaller sets of media data. However, the findings
obtained from the analysis are difficult to generalise beyond the studied cases. Due to this reason,
analysts interested in media discourses are increasingly adopting the tools and methods of Corpus
Linguistics (CL), which provides the study to be more empirical (O’Halloran, 2010).
An alternative to these traditional qualitative analyses is Corpus-assisted Discourse Studies
(CADS), which offers a convenient method through corpus investigation, combining close reading
with statistical analysis in the investigation of media discourse. CADS approach allows for
quantitative results produced by corpus tools to be studied in more depth using the qualitative
discourse-analytic studies. The combination of the quantitative corpus techniques with more
qualitative methods has been identified as effective methodological synergies (Partington, 2006;
56
Siti Aeisha Joharry & Syamimi Turiman
Baker et al., 2008) to reveal language patterns and representations in the media when compared
with a quantitative or qualitative analysis alone.
2.2
Corpus-assisted Stylistics: A Practical Approach to Examine Rhetoric in Online
Columns/ Opinion Pieces
The research on investigating writing style or ‘stylistics’ have originally been literary, including
those that incorporate corpus such as Culpeper’s (2002) analysis of Shakepeare’s characters in
Romeo and Juliet and Mahlberg’s (2013) work on Dickens, to name a few. However, more corpus
studies have examined stylistic features such as irony, metaphor and word-play to be found in nonliterary discourse such as those in newspaper writing (e.g. Semino & Short, 2004; Bednarek, 2006).
Two similar characteristics of corpus stylistic studies and CADS are that both approaches
are usually (if not, mostly) contrastive and that methods/techniques are a combination of various
linguistic analyses and definitely employ use of computational software like WordSmith Tools,
AntConc, WMatrix and others. More importantly, Partington et al. (2013) assert that the
incorporation of statistical corpus methods to studies in stylistics are essential as to highlight
“particular distinctive tendencies within an author’s – or groups of authors’ – style which may
warrant further scrutiny” (p.168). A number of studies examine metadiscoursal features as a
rhetorical distinction of opinion pieces as a genre like Fu and Hyland (2014), and Mengyu and
Hajar (2019). Noorian and Biria (2010) more specifically, compared opinion pieces written by
American and Iranian columnists and discovered that the choice of metadiscourse markers are also
culturally driven.
Hyland (2005) asserts that metadiscoursal features are facilitating tools on social
communication, which in turn reveal the ways that writers project themselves into their
argumentation in order to control their interactive intentions and signal their perspectives and
commitments (p.14). Hyland’s (2005) interpersonal model of metadiscourse includes an array of
features which assist writers to relate a text to its context, and enables writers to control the level
of personality in a text to establish a suitable relationship with their audience. The model
recognizes that metadiscourse is comprised of the two dimensions of interaction, as can be seen in
Table 1 below:
57
Corpus Stylistic Analysis of Malaysian Online Columnists
Table 1: Interpersonal model of metadiscourse (Hyland, 2005, p49)
Category
Function
Examples
Interactive
Help to guide the reader
through the text
Transitions
express relations between
main clauses
in addition; but;
thus; and
Frame
markers
refer to discourse acts,
sequences or stages
finally; to conclude;
my purpose is
Endophoric
markers
refer to information in other
parts of the text
noted above; see
Fig.; in section 2
Evidentials
refer to information from
other texts
according to X; Z
states
Code glosses
elaborate propositional
meanings
namely; e.g.; such
as; in other words
Interactional
Involve the reader in the
text
Hedges
withhold commitment and
open dialogue
might; perhaps;
possible; about
Boosters
emphasize certainty or close
dialogue
in fact; definitely; it
is clear that
Attitude
markers
express writer’s attitude to
proposition
unfortunately; I
agree; surprisingly
Selfmentions
explicit reference to
author(s)
I; we; my; me; our
Engagement
markers
explicitly build relationship
with readers
consider; note; you
can see that
These features provide a means for understanding the ways in which writers manage interaction
and explicate how different kinds of texts demonstrate the interactive process of a discourse.
Corpus-assisted stylistic studies examining metadiscourse features in online opinion pieces thus in
58
Siti Aeisha Joharry & Syamimi Turiman
turn, are interesting to reveal how writers demonstrate rhetorical strategies in order to represent
their understanding of reality as well as drawing on a sense of commonality with their readers.
The purpose of this paper is to then explore stylistic differences between two local
journalistic commentary writing and how combining corpus and stylistic approaches to media
discourse is feasible in eliciting salient metadiscoursal features amongst columnists. In so doing,
we examined the language used in two Malaysian English language online columns. More
specifically, corpus techniques are employed to elicit differences among the two columnists’
writing style. We firstly describe the corpus and methodology below, then go on to analyse the
texts in the subsequent section.
3.
Methodology
This study is underpinned by the corpus linguistics methodology (see Tognini-Bonelli, 2001) and
in turn, the first step is to compile the corpus. This corpus consists of articles retrieved from the
columnist section of two widely readable mainstream Malaysian online newspapers: The Star
Online and News Straits Times. It is important to note here that these online news portals are
referred to as the online edition of their typical broadsheets. As Rogers (2019) rightfully points
out, “[t]he online versions of these broadsheets, however, are more tabloid-like in format; they
have flashier headlines, attention-grabbing colour, and more graphics than the print editions”.
Therefore, the caveat here is to not treat the online portals as their original broadsheet
equivalents since it is undeniable that this form of journalism “seems to be to capture readers'
attention with tabloid techniques instead of relying on a more in-depth, broadsheet-like, serious
tone” (Rogers, 2019), and therefore, would be more interesting to study.
In order to select columnists that would be appropriate for this study, a quick search was
done by clicking on the ‘Columnists’ tab of the respective online news portals to find writers with
the most articles. This eventually came down to Syahredzan Johan1 (a total of 132 articles at the
time of collection) for The Star Online and John Teo (259 articles) for the NST. The articles were
collected from the year the columnists began their writing career, which was the year 2013 for
John Teo and 2014 for Syahredzan Johan until September 2019. Although both writers did not
necessarily began writing for their columns during the same time, it could be argued that both had
1
It was later discovered that Syahredzan Johan is no longer part of the columnist section in The Star Online, but his
writings should still be applicable for study.
59
Corpus Stylistic Analysis of Malaysian Online Columnists
almost similar years of experience for writing a column in a national newspaper agency and are
therefore considered to be professional writers/columnists in their own right. Table 2 shows the
two corpora in detail:
Table 2: Corpus description
Columnist
Words
Texts
Source
Syahredzan
90,721
132
The Star Online
John Teo
194,815
259
News Straits Times
(NST)
Using WordSmith Tools 6.0 (Scott, 2012), the two corpora are firstly compared to each other by
using the ‘compare two wordlists’ feature. This corpus technique basically helps make stylistic
comparisons. In other words, by comparing the separate wordlists, we are able to identify
similar/different words used between the two writers in their columns. The procedure compares
all the words in both lists and will report on all those which appear significantly more often in one
than the other, including those which appear more than the minimum number of times in one even
if they do not appear at all in the other (Scott, 2012). In this study, we restricted the words to those
occurring with a minimum of 10 times in at least 5 texts using log-likelihood. By doing this,
keywords of one columnist means that these words are statistically more significant than its
occurrence in the other column.
Following this, shared keywords in both corpora will be analysed further since salient
words that are used by both columnists would be more interesting for comparing purposes.
Collocational analysis, i.e. investigation of co-occurring words surrounding selected keywords is
then conducted by specifying associations within a span of 5 words to the right and left of a word,
occurring at least 1 time, and with an MI-score value of more than 3. As mentioned in Hunston
(2002), MI measures the strength of association between a word and its collocate. Having
identified salient keywords and their collocates, further qualitative analysis is then carried out by
investigating metadiscoursal features following Hyland (2004; 2005), and examination of the first
person pronoun ‘I’ (McNair, 2008).
60
Siti Aeisha Joharry & Syamimi Turiman
4.
Findings and Discussion
In order to examine stylistic differences between the two columnists, the two wordlists are firstly
compared using the ‘compare two wordlists’ function in WordSmith Tools 6.0. Table 3 presents
top 50 key words that appear significantly more often in Syahredzan than in John Teo’s. Initial
observation shows that a number of words relates to law/policy (e.g. act, law(s), court) and
government/public (e.g. constitution, parliament), which would describe the topics that are more
commonly written by Syahredzan compared to John Teo. Some words are not found in John Teo’s
columns (e.g. seditious, offences, liberties, and mentri), and while other topic-related words such
as ‘act, law/laws, court, and constitution’ were found, they appeared only few compared to being
more frequent in Syahredzan’s writings. This is probably because Syahredzan is a lawyer by
profession and his writings are more suggestive of the law/political genre. These initial findings
from Table 3 reveal a clear distinction of what topics may be found in one corpus relative to another
- in this case, genre plays a significant role in projecting content words that are more topic-related
to the discipline and thus, would be used to describe Syahredzan’s style of writing as more political
over John’s.
Interestingly, some words that are not topic-related are found to be significantly more
frequent in Syahrezdan compared to John Teo’s, namely personal pronouns (you, we, your, I),
modal verbs (would, must, should), deictic marker these, existential/action verbs (is, are, do),
conjunctions (or, also), article the, and the negated form not. This means that Syahredzan uses
these words more often than John Teo would and therefore, firstly suggests how different their
writing could be. In contrast to content words that would indicate the aboutness of the texts
mentioned earlier, functional words describe more of the writers’ style (or rhetorical strategies)
and in turn, are examined further in 4.1 using Hyland’s (2005) metadiscoursal features.
61
Corpus Stylistic Analysis of Malaysian Online Columnists
Table 3: Top 50 words that are significantly more frequent in Syahredzan
compared to John Teo
N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Key word
act
you
law
person
court
laws
or
information
submission
humble
constitution
syahredzan
myt
johan
freedom
the
monday
we
offence
am
not
speech
sedition
article
expression
is
parliament
your
are
would
police
security
section
religion
bill
must
people
these
according
rights
legal
criminal
right
do
muslims
freq. in Syahredzan
Johan_files
freq. in John
Teo_files
Keyness
335
238
304
194
222
177
739
167
133
133
184
111
111
111
161
7036
107
631
105
120
983
94
88
96
83
1689
151
89
771
282
89
98
60
79
101
348
261
267
63
156
86
62
173
240
87
56
25
73
15
33
15
606
13
1
2
27
0
0
0
22
11988
1
564
2
10
1135
4
2
8
3
2430
51
8
895
204
14
21
1
10
25
312
198
206
4
81
20
6
107
191
23
490.52
400.07
382.93
348.68
338.15
312.30
308.20
299.73
294.10
285.83
281.39
254.63
254.63
254.63
251.74
249.29
234.86
226.40
222.51
212.42
198.35
185.26
184.20
169.92
166.65
157.47
157.11
155.02
152.99
141.75
132.96
129.94
128.18
126.29
125.24
123.91
122.49
122.18
117.25
115.37
109.88
106.20
106.19
104.67
104.33
62
Siti Aeisha Joharry & Syamimi Turiman
46
47
48
49
50
4.1
dewan
seditious
media
offences
liberties
56
44
140
42
41
4
0
75
0
0
102.11
100.91
100.38
96.33
94.03
Comparison of WordList Feature
Given that the comparison feature includes all significant key words occurring at least once in a
wordlist regardless of the other list, we specified only shared keywords between both corpora to
be analysed further (see Appendix). The shared keywords were then further categorised based on
Hyland’s (2005) interpersonal model of metadiscourse, as seen in Table 4 below. This resulted in
the words appearing in both Syahredzan and John’s columns that are more comparable for analysis
–highlighting that the values in brackets represent words occurring statistically more significant in
Syahredzan than in John Teo’s. It should be noted that the identification of metadiscoursal features
based on individual lexical items, especially interactive ones, may be misleading without looking
at the context further. Therefore, the list of examples below serves only as a description of possible
metadiscourse found in both columnists’ writing.
Table 4: Metadiscoursal features used by both columnists
Category
Interactive
Transition
Frame markers
Endophoric
markers
Evidentials
Code glosses
Interactional
Hedges
Boosters
Attitude markers
Engagement
markers
Self-mention
Examples
also (91.27), however (27.50)
who* (60.62), found* (29.67)
who (60.62), say (40.82), found (29.67), claim
(33.77)
should (90.63), would (141.75)
want (32.17)
we* (226.40), you (400.07), people (122.49),
muslims (104.33), citizens (35.78), society (58.15),
rakyat (41.43),
I (88.62), we (226.40)
* These examples may have two functions within the same category and therefore is listed twice
63
Corpus Stylistic Analysis of Malaysian Online Columnists
As can be seen in Table 4, interactive features only include transitions (also, however), endophoric
markers (who, found) and evidentials (say, claim). It is important to point out here that ‘who’ and
‘found’ may also function as evidentials where writers refer to external sources that help them
build their stance, for instance He cites the example one of the rightly guided Caliphs, Umar alKhattab, who suspended the hudud crime of theft during a year of famine (Syahredzan) and The
Court of Appeal found that Section 3(3) did not meet the test of proportionality (Syahredzan).
Frame markers and code glosses are not found to be frequent in both columnists’ writing and this
may signal a lesser preference to do discoursal acts, sequences (finally; to conclude) or elaborate
on propositional meanings (namely; e.g.) (Hyland, 2005). However, as mentioned earlier, further
examination of the concordance lines is needed to discern how interactive metadiscourse is used
in both writings.
In terms of interactional resources, Malaysian online columnists employed engagement
markers the most, through the use of explicitly addressing readers - noticeably by lexical items
such as pronouns we and you, and those signalling groups of people (people, muslims, citizens,
society, and rakyat). The columnists attempt to create a sense of unity or solidarity through
addressing the readers as ‘people’ (e.g. A Malaysian prime minister is accountable to his cabinet
colleagues, to Parliament, to components of his ruling coalition, to states, to the rulers and
ultimately, to the people come time for a general election - John Teo); ‘muslims’ (e.g. And Malays
and Muslims are made to look irrational and extreme, not to mention illogical - Syahredzan; As
Malaysian Muslims celebrate overcoming personal trials and tribulations over Ramadhan… John Teo); and ‘citizens’ (e.g. The State's business is not to ensure that citizens are not offended Syahredzan). Another common group reference is to the people of Malaysia often referred to as
‘rakyat’, for instance, Instead of relying on the Opposition parties to balance the Government, the
rakyat can be at the frontline keeping the Government in check (Syahredzan). These examples of
collective references reveal how columnists position themselves interactionally, where O’Keeffe
(2006) describes as reflecting a collective identity.
However, concordance lines for ‘rakyat’ in John Teo’s columns were not used for the same
purpose in Syahredzan. Instead, John Teo made reference to political parties like Pakatan Rakyat,
Parti Rakyat Sarawak, Parti Bersatu Rakyat Sabah and others, indicating a specific audience as
presented in Figure 1:
64
Siti Aeisha Joharry & Syamimi Turiman
Figure 1: ‘Rakyat’ in John Teo
For the engagement marker ‘we’, the caveat here is to recognise the use of the personal plural
pronoun as an inclusive we, which could be used in contexts that refer to shared knowledge or
form of solidarity, and in turn may also act as self-mentions. Inclusive we as a self-mention, also
reflects a solidarity marker in which O’Keeffe (2006: p137) states, functions as “to appropriate
speaking authority on behalf of audience” and therefore inadvertently engages readers to the text.
This can be seen in the examples below. The first person pronoun ‘I’ on the other hand, would be
explained later in the next section along with discussion on the use of ‘I’ as a trait of the columnist
following McNair (2008).
Examples of inclusive-we
1)
For us in Malaysia, a far more effective use of any leverage and influence we may
have over the course Myanmar takes will not be one of disapproving censure, but rather of
pointing it in the right direction — perhaps by way of the power of our own example, for
what it may be worth (John Teo)
65
Corpus Stylistic Analysis of Malaysian Online Columnists
2)
Taking offence is normal. It is a human reaction. We have different degrees of
tolerance (Syahredzan)
Fu and Hyland (2014) noted that ‘you’ is a highly interactive feature which grabs the reader’s
attention and invites a direct involvement in the unfolding argument (p.10). Interestingly,
Syahredzan was found to use ‘you’ significantly higher (400.07), suggesting that he makes more
involvement of the readers in the process of communication, making them (i.e. readers) his
‘partner’ in the process (e.g. Meanwhile, Article 9 provides that citizens, like you, have the right
to freedom of movement). On the contrary, John Teo makes less instances of ‘you’, and when he
does so, they resemble expressions like be careful what you wish for, you bet we do, …if you will.
While this paper does not make comparisons with other personal deixes like ‘they, them, their’, it
would be interesting to look at how columnists view themselves and their readers in what O’Keeffe
(2006, p. 139) argues as the “‘otherness’ [that] is constructed through pronoun choice”.
The columnists also employed hedges frequently to ‘withhold commitment to a claim’
(Hyland, 2005, p. 52). In the data, it was found that both columnists used should and would
relatively frequent, allowing for subjectivity of a proposition to be presented as an opinion rather
than a fact. Examples include proposing something to be a certain way (e.g. This is how it should
be -Syahredzan; There should be nothing to prevent Sarawak asserting all its political rights
overnight if it chooses –John Teo), suggesting something as the opposite (e.g. A caretaker
government should not, for example, do anything which may bind the new government Syahredzan; That said, the lesson of Umno’s downfall should not be lost on PBB –John Teo); and
asking a rhetorical question (e.g. Should you just write off the debt? -Syahredzan; Should it be
making a stand publicly about probably the single biggest national scandal ever and which
unseated an incumbent national government for the first time? –John Teo). These then describe
the way in which columnists present their personal views and perspectives towards a certain issue,
particularly by minimising their imposition about a claim. This concurs with the findings of
Noorian and Biria (2014) who mentioned that hedging can be considered as ‘an important
characteristic of professional writing’. Thus, hedges appear to be an essential element in expressing
persuasiveness in online opinion columns.
While there is little variety in the use of attitude markers in both columnists’ opinion pieces,
writers are found to employ ‘want’ to demonstrate a sense of volition as well as indicate their
66
Siti Aeisha Joharry & Syamimi Turiman
affective attitude towards certain propositions. Interestingly, samples of the use of want in both
columnists’ writings can be classified into a number of categories. The first one is to denote
volition from the perspective of the writer (e.g. I am in danger of burning out so I need to pace
myself as I want to be in this for the long run -Syahredzan). Volition is also expressed in terms of
the inclusive ‘we’ (e.g. Malaysians want a more open and transparent and accountable
government. We want to know what the government is doing -Syahredzan; but yet we somehow,
almost inexplicably, always want to cling on to…- John Teo), as well as to position the volition to
others (e.g. They do not want a cooperation with individuals which they do not like - Syahredzan;
the new chief minister will have his own leadership style and will want to stamp his own imprint,
- John Teo). Another type is where ‘want’ is used in a hypothetical ‘if’-statement such as but if we
truly want to empower the people, this is the way forward (Syahredzan).
Since opinion pieces mainly function to convey the columnists’ stance on a topic and to
recruit readers into agreeing to their stance, it is worth noting that the attitude marker ‘want’ appear
in the corpora mostly to highlight what writers feel and how they feel about what others may
think/feel. One of the differences in the writing styles of the two columnists is that Syahredzan
demonstrates a wider use of volition by employing ‘want’ when expressing affective attitude
towards propositions. For instance, he expresses his hope about his personal journey as a lawyer
with his readers (example above), involving the reader via engagement marker ‘you’ (e.g. You are
still free to say whatever you want, even about the person who blocked you) and promoting
solidarity through use of the inclusive ‘we’ (example above), as well as presenting something in a
hypothetical way via the ‘if’-clause (e.g. What if he does not want to resign?).
Fu and Hyland (2014, p. 22) assert that ‘such strong expressions of attitude insinuate a
shared response to situations, implying agreement on understanding the world and bringing readers
around to the author’s uncontroversial conclusions’, as seen in the following examples:
3)
However, if you want to make institutional reforms or change the law, it must be
through the legislative process, in which our elected representatives represent the larger
populace [...] To say that you do not want to choose a candidate because there are no real
choices is an argument that does not hold water - Syahredzan
67
Corpus Stylistic Analysis of Malaysian Online Columnists
Interestingly, John seems to express this sense of volition through instances of hedging, such as
shown below:
4)
Malaysians, like Americans, always conscious of how their respective carefully
ordered societies may potentially be rent asunder by unforeseen events or careless individual
remarks, however inconsequential they may appear if taken in isolation, may want to take to
heart the socially redeeming lesson that Charleston offers us - John Teo
From these examples, it is observed that attitude markers did not only signal the columnists’
attitude towards certain propositions, it can also be seen that Syahredzan performs these actions –
specifically through use of ‘want’ – more frequently and without “beating around the bush”. Also,
use of attitude markers like ‘want’ in the instances shown above create an informal tone and thus,
making Syahredzan’s writings more personal, “creating a bond with the readers and securing the
readers’ agreements with their [Syahredzan] arguments” (Fu & Hyland, 2014, p22). The last
metadiscoursal feature which is self-mention will be examined more closely in the ensuing subsection.
4.2
The Use of ‘I’
McNair (2008, p. 109) stresses that “[a]s opposed to the detached objectivity of the reporter or
correspondent, the columnist stresses the ‘I’”. As can be seen in Table A1 in Appendix, the use of
‘I’ is over-used in Syahredzan (18%) as opposed to John Teo’s (6%). Collocational analyses of the
use of ‘I’ with the immediate word to the right (I + 1R) in both columns are then further examined.
This means patterns would indicate how writers used the first person pronoun to show what they
think/feel about something and most importantly, their opinions about a piece of news.
Table 5: Collocates for ‘I’ in both corpora
Columnist
Syahredzan
John Teo
Collocates for ‘I’
myself (8.71), wrote (7.96), am (5.68), feel (5.67), love (5.43), know
(5.25), did (4.47), also (3.87), was (3.55), do (3.54), and have (3.36)
am (10.64), think (9.21), would (5.7), can (4.44), had (4.35), only
(4.08), was (3.99), and have (3.3)
68
Siti Aeisha Joharry & Syamimi Turiman
Based on the collocates appearing immediately after the self-mention ‘I’ (Table 5), it is found that
Syahredzan has more variety in the use of the personal pronoun than his counterpart. One
interesting finding is the different use of verbs to describe mental processes with the self-mention
‘I’, i.e. Syahredzan uses more emotive verbs like I feel/love and therefore appears to be more
personal compared to John. Syahredzan also uses the reflexive pronoun ‘myself’ after ‘I’ (I myself
have been guilty of this; I will tell myself that we can rise above the prejudice and suspicion if only
we focus on our similarities instead of our differences), which in turn increases/intensifies the
personalization of his writing. It is also found that Syahredzan uses the cognitive verb ‘know’ (see
Figure 2) whereas John uses ‘I’ with cognitive verbs like ‘think’ when referring to personal
opinions (Figure 3). This implies that Syahredzan is more assertive in his claims while John
remains cautious.
Figure 2: ‘I know’ in Syahredzan
Figure 3: ‘I think’ in John
Similar use of self-mention ‘I’s in both columnists’ writings are highlighted in bold above,
including existential use of the self-mention with patterns like I am, I was and I have. However,
further investigation of concordance lines reveal that they are used differently between the writers.
69
Corpus Stylistic Analysis of Malaysian Online Columnists
Although 3 out of 9 of John’s examples of ‘I am’ are actually quotes from other people, he
expresses his conviction and confidence when he uses I am + so sold/even seriously contemplating
on talks about UTC. He also shows his confidence in relation to his belief of the society as a form
of solidarity in
5)
Therefore, if we believe that our society has progressed and grown in sophistication
over the years (which I am confident means most of us), we must necessarily place greater
trust in our own capacity to keep our own society progressing, going forward.
Another use of the self-mention phrase is used when John reflects on the past:
6)
I am reminded of almost similar conditions now prevailing in our country obtaining
a generation ago when the so-called “people power” revolution of 1986 overthrew then
Philippine president Ferdinand Marcos.
Other forms of ‘I am’ occur in expressions that suggest hedging like “…as far as I am concerned”,
and “I am sure that…”. Meanwhile, Syahredzan makes more use of the ‘I am’ phrase as shown in
Figure 4 below. This variation of use, especially those expressing attitudinal stance (e.g. I am lucky
enough; I am clueless about; I am honoured/still passionate about; I am not lazy; I am serious, no
kidding and others) suggests that his writing is more personalised.
70
Siti Aeisha Joharry & Syamimi Turiman
Figure 4: ‘I am’ in Syahredzan
As regards the use of existential phrase ‘I was’, it was found that Syahredzan makes more mention
of his emotional state of being, which was not found in John. This adds more to Syahredzan more
personalised views than his counterpart (examples 7 – 12).
7)
I was angry with those who had committed such a heinous act on innocent civilians.
And I feared for the backlash on Muslims living in Britain at that time.
8)
I was more reckless, brash and less matured, so I had to learn the hard way through
these mistakes.
9)
As I watched the scene of mayhem from my television screen in Cardiff that day,
emotions of sadness, anger and fear swept over me. I was sad for the country I lived in.
10)
I recently decided to trim my Facebook friends list. I was foolish enough before to
approve almost all friend requests that came my way without any sort of filtering process.
11)
Twitter was still new to Malaysians at that time and I was even lucky enough to
participate in what was probably the first Malaysian ‘tweetup’ (tweetup = Twitter meet up),
way before tweetups were ‘uncool’ as they are now.
12)
I was also worried about the overnight stay as there was nothing to stop
troublemakers from disrupting the rally in the dead of night.
71
Corpus Stylistic Analysis of Malaysian Online Columnists
In terms of the final shared use of self-mention ‘I’ among both columnists, Syahredzan (again)
displays a more emotive stance with the use of passive form ‘I have been’ (e.g. I have (also) been
lucky/blessed/always been passionate). Interestingly, he also makes mention of his actions in the
past, and to which he reflects on in his writings (examples 13 – 15).
13)
When I see some senior lawyers on their feet, articulately submitting to the court
with eloquence and a touch of muted passion, seamlessly weaving facts and law with little
effort, I know that I have a long, long way to go and much to learn.
14)
I have learned much over the past seven years.
15)
Yes, I have made many, many mistakes, both outside court and in court.
Other instances reflect Syahredzan’s personal relationship with his close friends (e.g. I have known
this group/them/these friends) and further add to his personalised writing style. John on the other
hand, only used the ‘I have’ phrase in the following two instances. Interestingly, it can be seen that
he constructs this phrase as an embedded clause and in turn, adds to the description of hedging in
his writing.
16)
The early years of Merdeka right up to Malaysia Day in 1963 are, of course, years
I have little to no personal memory of, except for that etched in my young mind of being
caught literally in the crossfire in the Limbang revolt of 1962 as our family hid in the concrete
bathroom to dodge live bullets fired into our wooden, stilted government bungalow.
17)
Because, since giving up on a regular 9 to 5 job, what I have discovered to be one
of the more liberating perks is not to be held slave to the electronically-inspired call of duty.
5.
Conclusion and Recommendations
To reflect on the digital news reports by Reuters (2018, 2019), it is inevitable that from today
onwards, more news will be read online and this includes commentary writing. Although we have
acknowledged the discrepancy of online versions and their broadsheets, we argue that the wider
readership of columns online rather than its broadsheet counterpart show that columnists have
more influence if any, to be more persuasive in the digital realm. As Hall (2008) points out,
“community journalisms [such as columnists], which before the web were regarded by the larger
72
Siti Aeisha Joharry & Syamimi Turiman
news organisations as marginal and unprofitable, […] have helped to show journalism how to
make news a fully interactive process” (p. 211) and that the “personality of particular writers,
expressed in style and content, became a way of attracting readers to a publication” (McNair, 2008,
p. 109).
This paper for the most part, has attempted to show this by investigating Malaysian
columnist English writing that focused on the examination of metadiscoursal features as rhetorical
strategies, via corpus linguistics. Two columnists from The Star Online and News Straits Times
were analysed to discern stylistic differences in their writings. Findings revealed that while
differences show topic-related words are nothing interesting to the writers, shared functional words
on the other hand reveal some distinctions. In brief, Syahredzan was depicted as more personal in
his writings compared to John Teo, mainly given his more frequent and various use of interactive
and interactional metadiscourse features. The presence of metadiscourse in the corpus of
Malaysian opinion columns was found to be consistent with the findings of Dafouz-Milne (2008)
and Noorian and Biria (2014), in that the construction of persuasion in the newspaper opinion
columns can be achieved through the use of metadiscoursal features.
The comparisons between the two experienced columnists have revealed distinctive styles
and their stylistic projection of newsworthiness. Bednarek (2015) notes, newsworthiness is
“treated as a quality of news media talk and text, and news values are conceptualised as the
‘newsworthy’ aspects of actors and events as constructed through linguistic resources (p28).” In
particular, the findings described how the columnists would attract readers to their articles,
specifically when Hyland’s (2005) metadiscoursal features were employed. As argued in the paper,
columns are generally another way of reporting news, and how they are written would therefore,
function to connect or engage with readers. The stylistic differences between the two columnists
in this study therefore contribute to the description of journalistic discourse. As McNair (2008)
puts it, the columnist stresses the ‘I’, as opposed to the detached objectivity of the
reporter/correspondent. The use of the personal pronoun ‘I’ has been found to characterise online
columns by producing a sense of solidarity with readers and in turn, contributing to reader-writer
interaction/visibility. As noted by Coward (2013), “columns not only use personal pronouns and
even personal anecdotes, they require them. The writing is all about communicating a distinctive
personality both in content and style, the more distinctive, the better” (p. 35).
73
Corpus Stylistic Analysis of Malaysian Online Columnists
Despite the differences in style choices, both online columnists commonly employ
interactional metadiscourse features as a means to organize the discourse; to help them create
meanings and negotiate their ideas with their readers. It can be suggested that the choices that
columnists make in their writing may determine the effect on their readers. While Syahredzan may
elicit a more personalized and assertive stance (e.g. I feel/love/know), John uses the ‘I’ with the
cognitive verb ‘think’ and modal verbs like ‘would’ and ‘can’, indicating scepticisms or
reservations. Unique to news reporting, columnists have a significant role in reporting news, as
McNair (2008) highlights, they have the added flair to present and analyse the news via their
personal lens. Pape and Featherstone (2006) assert that “personal columns and the views expressed
in them are part of a newspaper’s personality. Certainly, they are one of the very few places where
a dialogue is established between reader and writer and as such, they have an important role in
establishing the personality and soul of a paper” (p. 89). This characterises the online column
discourse, in which the readers have a special relationship with the columnists. Readers look to
specific columnists for a distinct style, point of view and informed opinion. In other words, it is
the columnist, not the newspaper that attracts the readers.
Thus, online columns may reflect attitudes and viewpoints of a society and in turn,
influence our (readers’) reaction to various issues, since they are written by independent experts
from various vantage points in the society. This is notwithstanding that more recent views on
columnists have declined – mostly due to the digital change/development of things that creates an
increase of subjectivity, lesser objectivity in presenting the news. McNair notes that “[t]he rise of
the internet has sharpened debates about the mobilisation and measurement of trust and authority
in journalistic discourse” (2008, p. 113) and that fake news, for example may be a result of
declining rhetoric in journalism. In light of this, digitizing has also provided a platform for readers
to even respond to opinion pieces (consider the reader’s feedback at the bottom of the online article
–sometimes together with emojis/likert scale for feedback), resulting in what we call as an opinion
within an opinion. This in turn, creates a wealth of possible future research avenues. Additionally,
this study reveals ways to use corpus techniques for analysing news media/content (i.e. news,
articles, and column pieces), which may benefit those who wish to study media and journalism.
74
Siti Aeisha Joharry & Syamimi Turiman
Acknowledgements
We would like to thank the anonymous reviewers for their constructive comments/feedback that
have helped shape the final manuscript.
References
Baker, P., Gabrielatos, C. & McEnery, A. (2013). Discourse Analysis and Media Attitudes: The
representation of Islam in the British Press. Cambridge: Cambridge University Press.
Baker, P., Gabrielatos, C., Khosravinik, M., Krzyzanowski, M., McEnery, T., & Wodak, R. (2008).
A useful methodological synergy? Combining critical discourse analysis and corpus
linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse
and Society, 19(3), 273–306.
Bednarek, M. & Caple, H. (2014). Why do news values matter? Towards a new methodological
framework for analysing news discourse in Critical Discourse Analysis and beyond.
Discourse & Society, 25(2), 135–158.
Bernstein, A., & Galily, Y. (2008). Games and sets: Women, media and sport in Israel. Nashim A
Journal of Jewish Women’s Studies and Gender Issues, 15(1), 175–196.
Biber, D., Johansson, S., Leech, G., Conrad, S. & Finegan, E. (1999). Longman Grammar of
Spoken and Written English. Harlow, England: Pearson Education.
Bird, S. E.. (2011). Seeking the audience for news: Response, news talk, and everyday practices.
In: V. Nightingale (ed.). The Handbook of Media Audiences. Oxford: Blackwell Publishing
Ltd., 489–508.
Coward, R. (2013). Speaking personally: The rise of subjective and confessional journalism.
Hampshire: Palgrave Macmillan.
Culpeper, J. (2002). Computers, language and characterisation: An Analysis of six characters in
Romeo and Juliet. In: U. MelanderMarttala, C. Ostman and Merja Kyto (eds.), Conversation
in Life and in Literature: Papers from the ASLA Symposium, Association Suedoise de
Linguistique Appliquee (ASLA), 15. Universitetstryckeriet: Uppsala, 11–30.
Dafouz-Milne, E. (2008). The pragmatic role of textual and interpersonal metadiscourse markers
in the construction and attainment of persuasion: A cross-linguistic study of newspaper
discourse.
Journal
of
Pragmatics,
40
(1),
95-113.
https://doi.org/10.1016/j.pragma.2007.10.003
75
Corpus Stylistic Analysis of Malaysian Online Columnists
Djerf-Pierre, M., & Shehata, A. (2017). Still an agenda setter: Traditional news media and public
opinion during the transition from low to high choice media environments. Journal of
Communication, 67 (5), 733–757. https://doi.org/10.1111/jcom.12327
Fairclough, N. (1995). Media Discourse. London: Edward Arnold.
Fu, X., & Hyland, K. (2014). Interaction in two
journalistic
genres: A study of
interactional
metadiscourse. English Text Construction, 7(1), 122–144.
Khabbazi-Oskouei, L. (2013). Propositional or non-propositional, that is the question: A new
approach to analyzing ‘interpersonal metadiscourse’ in editorials. Journal of Pragmatics,
47(1), 93–107.
Hall, J. (2008). Online editions: newspapers and the ‘new’ news. In Franklin, B. (Ed.), Pulling
Newspapers Apart: Analysing Print Journalism (pp. 215–223). London: Routledge.
Hunston, S. (2002). Methods in corpus linguistics: Beyond the concordance line. In S. Hunston
(Ed.), Corpora in applied linguistics (pp. 67–95). Cambridge: Cambridge University Press.
Hyland, K. (2005). Stance and engagement: A model of interaction in academic discourse.
Discourse Studies, 7(2), 173–191.
Hyland, K. (2004). Disciplinary interactions: Metadiscourse in L2 postgraduate writing. Journal
of Second Language Writing, 13, 133–151.
Jaworska, S., & Krishnamurthy, R. (2012). On the F word: A corpus-based analysis of the media
representation of feminism in British and German press discourse, 1990–2009. Discourse &
Society, 23(4), 401–431.
Jaworska, S. (2016) Using a corpus assisted discourse studies (CADS) approach to investigate
constructions of identities in media reporting surrounding mega sport events: The case of the
London Olympics 2012. In: Lamond, I. R. and Platt, L. (Eds.) Critical Events Studies:
Approaches to Research. Leisure Studies in a Global Era. Palgrave Macmillan, Basingstoke,
pp.
149-174.
doi:
https://doi.org/10.1057/9781137523860_8.
Available
at
http://centaur.reading.ac.uk/43261/
Mahlberg, M. (2013). Corpus Stylistics and Dickens's Fiction. London: Routledge.
McEnery, T. & Hardie, A. (2012). Corpus Linguistics: Method, theory and practice. Cambridge:
Cambridge University Press.
McNair, B. (2008). I, Columnist. In Franklin, B. (Ed.), Pulling Newspapers Apart: Analysing Print
Journalism (pp. 106-114). London: Routledge.
76
Siti Aeisha Joharry & Syamimi Turiman
Mengyu, H.E., & Hajar, A.R. (2019). Comparing Engagement Markers in Economics Research
Articles and Opinion Pieces: A Corpus-based Study. GEMA Online® Journal of Language
Studies, 19 (2), 1-14. http://doi.org/10.17576/gema-2019-1902-01
Meikle, G. (2008). Naming and shaming: News satire and symbolic power. The Electronic Journal
of
Communication,
18(2,
3
&
4).
Available
at
http://www.cios.org/www/ejc/EJCPUBLIC/018/2/01847.html
Noorian, M, & Biria, R. (2014). Interpersonal Metadiscourse in Persuasive Journalism: A Study
of Texts by American and Iranian EFL Columnists. Journal of Modern Linguistics, 20, 64–
79.
O'Halloran, K. (2010). How to use corpus linguistics in the study of media discourse. In O'Keeffe,
Anne and McCarthy, Michael eds. The Routledge Handbook of Corpus Linguistics (pp. 563–
577). Abingdon: Routledge.
O’Keeffe, A. (2006). Invesitgating Media Discourse. London: Routledge.
O’Keeffe, A. (2011). Media and Discourse Analysis. In: Gee, J. & Handford M. (eds), The
Routledge Handbook of Discourse Analysis (pp. 441–454). London: Routledge.
Partington, A. (2006). The linguistics of laughter: A corpus-assisted study of laughter-talk.
London: Routledge.
Partington, A., Duguid, A. & Taylor, C. (2013). Patterns and Meanings in Discourse. Theory and
Practice in Corpus-Assisted Discourse Studies. Amsterdam: John Benjamins.
Pape, S. & Featherstone, S. (2006). Feature Writing: A Practical Introduction. London: SAGE
Publications.
Reese, S. D., Rutigliano, L., Hyun, K., & Jeong, J. (2007). Mapping the blogosphere: Professional
and citizen-based media in the global news arena. Journalism, 8(3), 235–261.
Reuters Institute Digital News Report. (2018). Oxford: Reuters Institute for the Study of
Journalism.
Reuters Institute Digital News Report. (2019). Oxford: Reuters Institute for the Study of
Journalism.
Rogers, T. (2019, July 15). Differences between Broadsheet and Tabloid Newspapers. Retrieved
from https://www.thoughtco.com/broadsheet-and-tabloid-newspapers-2074248
Scott, M. (2012). WordSmith Tools version 6, Liverpool: Lexical Analysis Software.
77
Corpus Stylistic Analysis of Malaysian Online Columnists
Semino, E. and Short, M. (2004). Corpus Stylistics: Speech, Writing and Thought. Presentation in
a Corpus of English Writing. London: Routledge.
Taylor, C. (2014) Investigating the representation of migrants in the UK and Italian press: A crosslinguistic corpus-assisted discourse analysis. International Journal of Corpus Linguistics, 19
(3), 368-400.
van Djik, T. (1993). Principles of critical discourse analysis. Discourse and Society, 4 (2), 249283.
Vaughan, E. & O’Keeffe, A. (2015). Corpus Analysis. In K. Tracy, C. Ilie & T. Sandel (Eds.). The
International Encyclopedia of Language and Social Interaction (pp. 252–268). Denver: John
Wiley & Sons.
Wodak, R. (2001). What CDA Is about—A Summary of Its History, Important Concepts and Its
Developments. In W. R., & M. Meyer (Eds.), Methods of Critical Discourse Analysis (pp.
1–13. London: Sage.
78
Siti Aeisha Joharry & Syamimi Turiman
Appendix
Table A1: Shared key words in both corpora
N
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Key word
the
of
is
not
by
have
are
this
was
or
we
there
they
who
must
also
should
these
would
people
under
do
were
I
right
without
why
cannot
rights
media
law
however
order
does
assembly
social
act
parliament
freq. in
Syahredza
n
Johan_files
7036
3036
1689
983
765
601
771
578
431
739
631
347
359
283
348
306
253
267
282
261
157
240
167
162
173
111
89
94
156
140
304
80
100
115
91
94
335
151
%
7.76
3.35
1.86
1.08
0.84
0.66
0.85
0.64
0.48
0.81
0.70
0.38
0.40
0.31
0.38
0.34
0.28
0.29
0.31
0.29
0.17
0.26
0.18
0.18
0.19
0.12
0.10
0.10
0.17
0.15
0.34
0.09
0.11
0.13
0.10
0.10
0.37
0.17
Texts
132
132
132
130
130
126
126
130
101
123
109
119
105
105
96
113
95
100
104
84
68
93
68
31
55
61
55
59
41
32
81
56
54
65
20
26
66
51
freq. in
John
Teo_file
s
RC. %
11988
6.15
5661
2.91
2430
1.25
1135
0.58
1061
0.54
897
0.46
895
0.46
767
0.39
670
0.34
606
0.31
564
0.29
414
0.21
394
0.20
319
0.16
312
0.16
300
0.15
226
0.12
206
0.11
204
0.10
198
0.10
194
0.10
191
0.10
185
0.09
110
0.06
107
0.05
88
0.05
84
0.04
83
0.04
81
0.04
75
0.04
73
0.04
73
0.04
71
0.04
70
0.04
65
0.03
61
0.03
56
0.03
51
0.03
Keynes
s
249.29
40.06
157.47
198.35
82.92
46.53
152.99
74.37
26.75
308.20
226.40
63.40
82.49
60.62
123.91
91.27
90.63
122.18
141.75
122.49
25.72
104.67
37.41
88.62
106.19
48.66
28.65
34.35
115.37
100.38
382.93
27.50
51.52
71.88
46.50
54.44
490.52
157.11
79
Corpus Stylistic Analysis of Malaysian Online Columnists
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
want
know
say
days
known
used
constitutional
citizens
action
religious
certain
society
human
court
comes
rakyat
constitution
young
you
bill
found
legislative
islamic
muslims
claim
freedom
security
legal
67
59
69
66
62
63
66
60
51
50
81
74
78
222
49
55
184
61
238
101
44
40
45
87
45
161
98
86
0.07
0.07
0.08
0.07
0.07
0.07
0.07
0.07
0.06
0.06
0.09
0.08
0.09
0.24
0.05
0.06
0.20
0.07
0.26
0.11
0.05
0.04
0.05
0.10
0.05
0.18
0.11
0.09
38
30
42
36
49
40
32
23
25
26
44
44
26
55
35
27
51
17
40
20
31
25
16
17
22
41
28
32
50
45
44
41
41
40
39
38
38
37
36
36
34
33
30
28
27
26
25
25
25
25
24
23
23
22
21
20
0.03
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
32.17
27.43
40.82
40.28
35.06
37.46
42.65
35.78
24.54
24.30
68.87
58.15
67.39
338.15
30.41
41.43
281.39
53.66
400.07
125.24
29.67
24.23
32.39
104.33
33.77
251.74
129.94
109.88
80