Open Journal of Modern Linguistics
2013. Vol.3, No.4, 295-304
Published Online December 2013 in SciRes (
Open Access 295
English and Malay Text Messages and What They Say about
Texts and Cultures
Ernisa Marzuki1, Catherine Walter2
1Universiti Malaysia Sarawak, Sarawak, Malaysia
2University of Oxford, Oxford, UK
Received September 11th, 2013; revised October 15th, 2013; accepted October 22nd, 2013
Copyright © 2013 Ernisa Marzuki, Catherine Walter. This is an open access article distributed under the Crea-
tive Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any me-
dium, provided the original work is properly cited.
This study of the pragmatics of cross-cultural text messages throws light on the evolution of new hybrid
forms of literacy and on the complex ways that culture is expressed and mediated in second language/
second culture contexts. An investigation was carried out into the pragmatics of apology in first-language
(L1) and second-language (L2) short messaging service text messages of adult Malay speakers who are
proficient users of English, living and studying in an English-speaking university environment; and into
L1 English users’ text apologies in the same context. Research questions included whether these profi-
cient L2 English users would perform differently from L1 English users in this high-stakes speech act,
and from their own L1 Malay use; and whether apologies in what has been called a hybrid medium would
differ from those previously studied in writing, in speech and in other electronic media. Twenty-six native
speakers of English and 26 native speakers of Malay responded via text messages to discourse completion
tests (DCTs) in L1; the DCTs represented either high or low levels of offence calling for apologies. The
Malay native speakers also responded to apology situations in L2 English. Data were coded using an
adapted version of Cohen and Olshtain’s (1981) coding scheme. Analysis of the messages sent by par-
ticipants revealed clear signs of a hybrid type of text that is differently conceptualised by the two commu-
nities. It also showed that the Malay users’ second language literacy was shaped in a complex way that
sometimes accommodated the second language/second culture and sometimes retained first language/first
culture values.
Keywords: Text Messages; Texting; Pragmatics; Apologies; Literacies; New Literacies
Cross-Linguistic Apologies by Text Message
Apologies attempt to rectify social discord caused by norm
violation (Scher & Darley, 1997). By apologising the speaker/
writer indicates acceptance of the violated norm, takes respon-
sibility for the violation and expresses regret for it (Aijmer,
1996), thereby attempting to remedy the offence caused (Tros-
borg, 1995). Apology attempts to preserve or restore the
hearer’s/reader’s face (Linnell, Porter, Stone, & Chen, 1992),
and is simultaneously face-threatening to the speaker/writer
(Brown & Levinson, 1978).
Given the importance of apology for social cohesion and the
potential for loss of face in the failure of high-stakes apologies,
it is not surprising that this speech act has received a great deal
of attention. Characteristics of apology have been shown to be
influenced by various factors, including the severity of the of-
fence (e.g., Grieve, 2010; Wouk, 2005), the interlocutor rela-
tionship (e.g. Mulamba, 2009; Shardakova, 2005), and gender
(e.g., Holmes, 1989; Hobbs, 2003). Performances and percep-
tions of apology have been extensively studied in the first lan-
guage (L1) communication of native speakers of a range of
varieties of English (NSEs), both adults (Grieve, 2010; Mu-
lamba, 2009; Kim, 2008; Kasanga & Lwanga-Lumu, 2007;
Sabate i Dalmau & Curelli i Gotor, 2007; Ancarno, 2005; Bha-
ruthram, 2003; Hobbs, 2003; Nakano, Miyasaka & Yamazaki,
2000; Linnell, Porter, Stone, & Chen, 1992; Sugimoto, 1997;
Olshtain, 1989) and children (Ely & Gleason, 2006; Kampf &
Blum-Kulka, 2007). In Blum-Kulka, House and Kasper’s
seminal 1989 work, they studied apologies in three varieties of
English, and also in French, German, Danish, Russian and He-
brew. They found little variation between languages in the use
of the five main pragmatic strategies for apology. (Note, how-
ever, that already in 1983 Olshtain and Cohen had found that,
unlike English apologies, Hebrew apologies were less likely to
include Offers of repair and Promises not to repeat offence than
English ones.)
Blum-Kulka et al. (1989) have called for more investigation
of apologies in non-Western cultures. This is in large part in
order to address the question of whether all human beings fol-
low a universal set of politeness rules, which has been debated
since Brown and Levinson’s (1978) original suggestion that
faceis a universal need which is addressed by politeness. Leech
(1983) proposes eight maxims of politeness; although he holds
these to be universal, he concedes that different cultures vary in
the extent to which they accept and/or use the maxims. Some
researchers, for example Wierzbicka (1991), maintain that since
each culture has its own unique norms, it is difficult to deter-
Open Access
mine universal characteristics of politeness. Since the Blum-
Kulka et al. (1989) study, there have been investigations of
apologies as performed in non-Western languages, for example
Sudanese Arabic (Nureddeen, 2008); Farsi (Afghari, 2007);
Ciluba (Mulamba, 2009), Setswana (Kasanga & Lwanga-
Lumu, 2007); Bahasa Indonesia (Wouk, 2005); Thai (Intachakra,
2004); Mandarin (Zhang, 2001); Japanese (Sugimoto, 1997);
and Korean (Kim, 2008; Park, Lee & Song, 2005; Yang, 2002).
Cross-Linguistic Apologies
Cross-linguistic apologies have also, and just as under-
standably, been investigated in several contexts. Whether po-
liteness rules are universal but are realised differently in differ-
ent cultures or cannot be considered as universal at all, it is
clear, as Mey (2001) notes, that pragmatic meanings are based
on societally-imposed conditions, and that at some level prag-
matic rules may differ from one society to another, resulting in
clashes of beliefs about what is polite and what is not. For this
reason cross-cultural apologies, as a potentially sensitive area
of politeness, have been the subject of several studies.
There have been a number of studies of various L1 - L2 (and
Dialect 1 - Dialect 2) pairings, for example English-French
(Cohen & Shively, 2007); English-Spanish (Cohen & Shively,
2007; Shively & Cohen, 2008); English-Russian (Shardakova,
2005); German-French (Warga & Scholmberger, 2007); Aus-
trian German-German German (Clyne, Fernandez, &Muhr,
2003); Swedish-German (Bohnke, 2001). However, the largest
number of cross-cultural apology studies have compared apolo-
gies by native speakers of English (NSEs) with those bynon-
native speakers of English (NNSEs). Some studies of NNSE
apologies have investigated a range of first languages (e.g.,
Linnell, Porter, Stone & Chen, 1992; Ancarno, 2005). Others
have focused on NNSEs from specific L1 backgrounds, for
example Catalan (Sabate iDalmau & Curelli iGotor, 2007);
German (Grieve, 2010); Korean (Yang, 2002); Japanese (Na-
kano, Miyasaka, & Yamazaki, 2000); Cantonese (Rose, 2000);
Mandarin (Chang, 2008); and Setswana (Kasanga & Lwanga-
Lumu, 2007). There is one investigation of Malay NNSE
apologies (Maros, 2006), in which participants completed writ-
ten DCTs to record what they would say in spoken English
situations; this study will be discussed below.
Malay “Cultural Scripts” and Apology
Previous studies have linked cultural concepts to perceptions
of politeness. For example, Chang (2008) investigated the per-
ceptions of Taiwanese NS of Mandarin and Australian NSEs
who listened to recordings of apologies and completed both a
questionnaire and a post-listening interview about levels of
politeness in apologies; Chang concluded that the Taiwanese
participants’ perceptions were linked to Chinese cultural con-
cepts. The Taiwanese participants in this study were argued to
base their perceptions on the Chinese cultural concepts of bu-
haoyisi (I feel embarrassed) and chengyi (sincerity).
Notably for the current discussion, Goddard (1996, 1997) has
carried out a careful analysis of a number of Malay “cultural
scripts”, based on his own data and on a re-analysis of data
reported by a range of Malay and non-Malay researchers of the
culture, using Wierzbicka’s (1996) Natural Semantic Metalan-
guage framework. This means that there is a solid foundation
for reflecting on possible cultural influences on Malay NNSE
apologies. Findings relevant to the present investigation of
apologies are that in the Malaysian culture:
It is important to be well-mannered or refined (halus) as
opposed to coarse (kasar). “A great deal of what it means to
be halus hinges on one’s speech” (Goddard, 1997: p. 186).
It is not appropriate to make explicit reference to one’s own
feelings or to the feelings of others; people are expected to
be sensitive enough to understand how other people are
feeling, and to be considerate towards their interlocutor’s
Rather than make direct reference to an offence, “[i]f I must
say something, it should be vague” (Goddard, 1997: p. 193).
Goddard gives as an example of an appropriate utterance “If
yesterday I did/said something [uncouth], I ask for pardon”
(Goddard, 1997: p. 193).
It is not polite to voice wishes about what other people
should do.
It is interesting to compare Goddard’s findings to the socio-
pragmatic components of apologies as categorised by the stan-
dard taxonomy introduced by Cohen and Olshtain in 1981 (later
expanded by Blum-Kulka et al., 1989):
1) Illocutionary force indicating device (IFID): an explicit
expression of apology, e.g. “sorry”, “forgive me”, “I apologise”
2) Explanation or account, e.g. “someone bumped into me”
3) Acknowledgment of responsibility, e.g. “I was careless”
4) Offer of repair, e.g. “I will buy you a new vase”
5) Promise not to repeat offence1
In the Malay context, two of the sample IFIDs would seem to
be problematic: “sorry” would appear to be excluded as refer-
ring to one’s own feelings; “I apologise” appears to be too di-
rect a reference to the offence. In this regard it is interesting
that Maros’s (2006) 27 Malay participants, who each completed
written English DCTs reporting what their oral response would
be in six situations, often used “Please forgive me” or “Excuse
me” in situations of serious offence or of offence to a high-
status interlocutor. “Please forgive me” comes closer to God-
dard’s (1997) “I ask for pardon” than does a straightforward
apology—as does “Excuse me” if its literal meaning rather than
its standard illocutionary force is taken into account.
The other four categories appear to make direct reference to
the offence. However, the L1 Malay avoidance of direct refer-
ence was not evident in the English responses of Maros’ (2006)
Malay participants, for whom Explanation or account was the
most frequent strategy, followed by Acceptance of responsibil-
ity. In Maros’ study, 27 Malay NNSEs’ written DCT apologies
in English were collected for six situations. Maros attributed
some of the characteristics of her participants’ responses to
transfer from Malay cultural norms, e.g. over-formality in low-
risk encounters. Maros’ participants had not lived outside Ma-
laysia, and she attributes the persistence of L1 pragmatic pat-
terns to the fact that their English had been spoken almost ex-
clusively with other Malaysians; this hypothesis will be tested
by the present study, where the Malay speakers were all living
and studying in the UK. Note too that Maros’ results can only
be suggestive. The written DCT format has been criticised, for
example by Woodfield, whose (2008) study of NSEs’ responses
to written DCTs found that written responses deviated from
1This is commonly called promise of forbearance in the literature (e.g.,
Blum-Kulka et al., 1989; Olshtain, 1989; Trosborg, 1987; Scher & Darley,
1997), but this does not correspond to the standard meaning of forbearance
in English. Therefore, we have chosen to use Bataineh and Bataineh’s (2008)
more transparent label.
Open Access 297
corresponding spoken responses. Further, in Maros’ study there
was no comparison NSE group, there were no comparison
DCTs completed in Malay, Olshtain’s (1983) taxonomy was
not applied in a conventional way, and the results are not re-
ported in detail.
Medium of Apology: Speech, Writing,
and E-Communication
A further variable taken into account in this study was the
medium in which the apology is performed. Most of the studies
cited above have dealt with either written or spoken apologies.
While there has been some recent interest in apologies con-
veyed electronically, we found only one study of the pragmatics
of text messages, and we argue that within the general category
of electronic communication text messages present specific
Some authors (Collot & Bellmore, 1996; Gains, 1999; Gi-
menez, 2000; Crystal, 2008a, 2008b) argue that the language
used in electronic means of communication such as emails,
instant messaging (IM) and Short Message Service (SMS) text
messages (henceforth text messages or texting) manifests a
sufficient number of distinctive characteristics for it to be con-
sidered a new medium alongside speech and writing. Electronic
communication, it is suggested, is a hybrid medium manifesting
some features of written language and some features of spoken
language. Some writers (e.g. Baron, 2000) argue that electronic
communication presages a move towards a unified standard
which will have more characteristics of spoken language than
of written language. While spoken and written language can be
seen as existing on a continuum, prototypical spoken language
happens in real time, is unplanned, is face to face, and reflects
an immediate interpersonal situation (Carter & McCarthy, 2006:
p. 164); while prototypical written language enjoys relative
permanence; lacks face-to-face contact, forcing both increased
explicitness (because of fewer shared contextual elements and
because of the lack of contemporaneous feedback) and the need
for proxies for suprasegmental and paralinguistic features; and
adheres to formal conventions (Crystal, 2005: pp. 149-151).
The degree to which the affordances of text messaging can be
said to correspond to these characterisations of speech or writ-
ing is problematic. Table 1 demonstrates the difficulty of situ-
ating text messages within the paradigm.
Indeed, within the general category of electronic communi-
cation, it can be argued that the affordances of text messages
differ substantially even from those of email and instant mes-
saging (IM). Differently from email, although there is no longer,
Table 1.
Comparing characteristics of spoken and written language with texting.
In real time Yes No Optionally
Unplanned Yes No Optionally
Face to face Yes No No
Immediate interpersonal situation Yes No Optionally
Explicitness No Yes Yes?
Formal conventions No Yes ?
as before, a strict limit on the number of characters in a text
message, text message format still tends to be restricted by the
size of the mobile/cellphone screen. Differently from IM, text
messages are sent without the knowledge of when the inter-
locutor will receive the message, and with the possibility but
not the firm expectation of an immediate response. These dif-
ferences lead to the hypothesis that the pragmatics of text mes-
saging may be substantially different from those of email or IM.
Some support for this hypothesis is provided by Baron (2004: p.
84), who found that US American university students consid-
ered that emails (in contrast with IM) should be “edited, punc-
tuated, spellchecked, and more formal”.
Studies of electronic communication thus far (e.g. the various
chapters in Danet & Herring, 2007; Ling & Baron, 2007), in-
cluding one study of text messaging in Malay (Badrul Redzuan,
2006) have mainly concentrated on the word-and sentence-level
characteristics of the medium or on the patterns of code-
switching and code-mixing in text messages. Less work has
focused on the discourse/pragmatic features of the medium, and
only Maros (2006) focuses on the pragmatic features of text
There have been a few studies of the pragmatics of apologies
by email. Hatipoğlu (2004) studied 126 emails: (a) one-to-one
(n = 56); (b) group from individual (n = 46); and (c) group from
official representative (n = 32) email apologies sent between
2002 and 2004, in English, in the context of a British university
department. The most marked difference was in the degree to
which apologies were intensified in the different conditions,
with intensified apologies containing words such as very, really
and so occurring much more often in the one-to-one condition
and only once in the group-from-official condition; group-
from-individual apologies split almost equally between intensi-
fied and non-intensified apologies.
Ancarno (2005) investigated a corpus of 66 letters and 86
emails to and from the editors of research journals written in
English by NSEs and NNSEs in three academic disciplines,
looking inter alia at the apologies in these emails. He found
that the writers of conventional letters tended to use “full”
apologies, with clear acknowledgment of responsibility, or to
use intensifiers in expressing their apologies, more than writers
of emails, who tended more towards “elliptical” apologies.
However, the email apologies tended to be for lower-stakes
offences, which may have been a confounding variable. An-
carno found no significant variations between NSEs and
NNSEs in the pragmatic characteristics of the apologies.
Research Questions
This study focuses on two areas: the potentially distinctive
characteristics of a representative speech act, apology, as real-
ised in text messages; and the potential for cross-cultural prag-
matic differences between the apologies of NSEs and Malay
NNSEs, as an under-researched linguistic/cultural pairing.
Three research questions emerge from the intersection of these
1) Can text message apologies in L1 and L2 English and L1
Malay be characterised in any sense as a hybrid medium be-
tween speaking and writing?
2) What if any are the pragmatic differences between text
message apologies in L1 English and those in L1 Malay?
3) Do highly proficient Malaysian speakers of English dem-
onstrate any Malay pragmatic characteristics in their English
Open Access
text messages?
26 native speakers of Malay (NSMs) (11 males, 15 females)
and 26 NSEs (10 males, 16 females) participated in the study.
All participants were undergraduate or postgraduate students in
universities in the United Kingdom. The participants ranged in
age between 19 and 41 years of age, with a mean age of 24.8
for English participants (SD: 5.32) and 25.5 for Malay partici-
pants (SD: 3.71). NSM participants had obtained at least a 6.5
on the International English Language Testing System
(IELTSTM) examination, 550 on the paper TOEFL®, 213 on the
computer-based TOEFL®, or A1 in GCSE O-level plus the
International Baccalaureate®. At the time of the study, all NSM
participants had been in the UK for no less than 2 months and
no more than 44 months (mean 16 months, SD: 10). Partici-
pants were selected through snowball sampling. Even though
the use of snowball sampling decreases the possibility of the
sample being representative of the whole population, this sam-
pling method was chosen to increase the possibility of recruit-
ing participants who met the criteria, and especially the criteria
for the NSM group.
Data was collected by Discourse Completion Task (DCT)
through text messages. This methodology gives a degree of
ecological validity to the data and is less vulnerable to the criti-
cisms of written DCTs for spoken situations voiced by Wood-
field (2008). The DCT scenarios were designed specifically for
text messaging. Since there was no precedent for texted apol-
ogy situations, the situations were designed by the researchers
based on previous apology studies such as those in the Cross-
Cultural Speech Act Realisation Project (Blum-Kulka et al.,
A preliminary study was carried out to identify 4 situations
to be used in the main research. Eleven situations were de-
signed and evaluated. The 11 situations were then given to 10
participants (five NSMs and five NSEs) who were not involved
in the main research. The participants were asked to rank the 11
situations from No Risk to Very High Risk of continuing or
exacerbating the offence. Risk was defined in relation to how
serious the norm violation was, and the consequent likelihood
that the receiver would be offended by an apology that they
perceived as insufficient. To avoid participants’ choosing a
middle ground such as Neutral or Unsure, a six-point Likert
scale was employed, with 1 representing a No Risk Situation
and 6 representing a Very High-Risk Situation. This yielded
four situations for the main study: the two with the highest
medians (of 5 and 4) for use as High Risk (HR) situations (for-
getting a meeting with a lecturer, forgetting to return a library
book for one’s supervisor) and the two with the lowest medians
(of 2 and 1) as Low Risk (LR) situations (forgetting to meet a
cousin, accidentally sending a text message to a wrong number).
These four situations were then edited in English and Malay
versions in order to make them more text message-like. The
receiver of the apology was specified as female in three of the
situations and unspecified in the fourth situation. This is be-
cause in Holmes’ (1989) study of 183 remedial interchanges by
New Zealanders in various contexts, male and female-directed
apologies were shown to differ: both men and women apolo-
gised more often to women. Note that the differences in social
status between the apologiser and the receiver contribute to the
risk level in all four cases, and we do not intend to investigate
social distance separately in this study.
A pilot study was conducted one month before the main
study. Four NSMs and four NSEs selected from the same sam-
ple pool as the main study responded to the four situations. As a
result of the pilot study, one of the situations was slightly edited
for clarity; and participants in the main study were given a
choice of receiving the prompts for the DCTs via phone calls or
text messages (although in the event no main study participants
opted for phone calls). The exact texts as sent to the main study
participants in English and Malay are given in Appendix A.
NSMs were required to respond to all four different situa-
tions, two in Malay and two in English. NSEs were only re-
quired to respond to two different situations, both in English.
Each NSE responded to one HR situation and one LR situation;
the situations were counterbalanced across the sample, so that
each situation received an approximately equal number of NSE
responses. For the NSM participants, the language used for
each situation was counterbalanced across the group to avoid an
effect of situation. The order of administration of the situations
was also counterbalanced to avoid practice and fatigue effects.
Participants were asked to respond to the situations by texting
the message which they would text if they were in the situations
described. The inclusion of textisms in the stimulus texts im-
plicitly encouraged the use of this convention.
On the agreed day, a researcher contacted the participants to
ensure that they were prepared to send and receive text mes-
sages. In cases where participants could not be reached or had
to postpone texting, new dates were set. Only after the partici-
pants declared that they fully understood what was expected
from them did the data collection began. The participants were
required to send a text message which stated “OK” to start the
data collection. Once this was received, the researcher sent the
first situation to the participant via text message. For ecological
validity, there was no time limit, as in an authentic context, a
person might take some time to compose a text message. How-
ever, the participants were reminded that they had to treat the
situation as urgent, thus discouraging them from taking too
long. In the three cases (two NSMs and one NSE) where par-
ticipants did not respond after 1.5 hours, the researcher re-sent
the situation, which successfully prompted all participants’
responses. Upon receiving the response to the first situation, the
researcher sent the second situation. In the case of the NSM
participants, the process was repeated until the fourth situation.
After the final response by the participant, the researcher sent a
text message indicating that the data collection was complete
and thanked the participant. The researcher offered compensa-
tion (GB £1.00 phone credit for each participant); all partici-
pants except two declined.
Sociopragmatic Coding System
A total of 156 responses were collected from participants,
with 26 High-Risk (HR) and 26 Low-Risk (LR) responses in
each of three categories:
Open Access 299
NSM-L1: L1 Malay responses from native speakers of Malay
NSM-L2: L2 English responses from native speakers of Malay
NSE-L1: English responses from native speakers of English
A coding scheme was developed based on Cohen and
Olshtain’s (1981) scheme as revised by Blum-Kulka et al.
(1989), comprising categorisation of component sociopragmatic
strategies, each with several different potential pragmalinguistic
realisations. Changes to the scheme were based on the patterns
which emerged from the actual responses from the participants.
As a result, ten types of pragmalinguistic realisations were
omitted from the Blum-Kulka et al. (1989) scheme and three
sociopragmatic strategy categories were added. The resulting
sociopragmatic categories are given below (new categories
emerging from the data are marked with asterisks), and the full
coding scheme is given in Appendix B.
1) Alerter: word(s) used to start the message and thus alert
the receiver to the message.
2) Illocutionary Force Indicating Device (IFID)
3) Taking/minimising/denying responsibility
4) Explanation or account: an explanation of how the event
was out of the apologiser’s control
5) Offer of repair
6) Promise not to repeat offence
7) Distracting from offence
8) Closing statement*: a type of complimentary close (e.g.
“best wishes”, “regards”).
9) Thanking receiver*
10) Concern for receiver/show of empathy*: a show of concern
or empathy for the receiver who may be affected by the offence
11) Others: other strategies which do not fit into to any of
the previous codes.
Initial coding involved dividing each text message into seg-
ments. Thus a text message like “sorry for the earlier msg. ac-
cidentally send [sic] it to the wrong person. sorry” would be
considered as containing three segments. Each segment was
then coded into its sociopragmatic moves and their pragmalin-
guistic realisations.
72 of the 156 messages (24 selected randomly from each of
the three categories) were coded by a second rater to test reli-
ability. The second rater was a university lecturer with a Mas-
ter’s degree in psycholinguistics who was teaching at a public
university in Malaysia at the time of the study. Inter-rater reli-
ability was 86.1%. Cases of disagreement were discussed and
consensus reached, and the remaining 74 messages were coded
on the basis of these discussions. This produced a rich data set,
of which only the most relevant results will be reported in detail
and discussed below.
Analysis of So ci opragmatic Str ategy Use
Strategy Differe nc e s between Groups and Languages
The frequency of each category of sociopragmatic strategy
was calculated. In the very rare cases where the same strategy
appeared more than once in the same text message it was only
counted once. Significant differences were calculated; the re-
sults are given in Tables 2 and 3.
High and Low Risk Categories
The situations had been designed to represent risk-level catego-
ries, High-Risk (HR) and Low-Risk (LR), and there were two
situations in each category. Fisher’s exact tests (because of
small numbers in cells) were carried out by risk category and
participant group to test the validity of the categories. There
were no significant differences between the two situations in
each risk level for any strategy except Offer of repair, where
the NSM group showed significant differences between both
the two HR situations and the two LR situations (p < 0.01) and
the NSE group showed significant differences between the two
situations for the LR category only (p < 0.01); and Alerters,
where the NSM group showed a significant difference between
the two LR situations only (p < 0.01). Therefore, these catego-
ries were excluded from analysis of differences between the
two risk levels.
A small number of differences were found between HR and
LR responses; all differences are reported here as determined
Table 2.
First-language apologies by English and Malay native speakers.
Sociopragmatic strategies in L1 Malay L1
English L1
Alerter 36 33
Illocutionary force indicating device 52 52
Taking/minimising responsibility 52 50
Explanation 1 2
Offer of repair 27 30
Promise not to repeat offence 1 2
Distracting strategy 12 6
*Closing statement 1 20
Thanking the receiver 4 1
Expressing concern/empathy 10 4
Note: *Fisher’s exact test (one-tailed) shows a significant difference at p < 0.05.
Table 3.
Apologies by native speakers of Malay (NSMs) in two languages.
Sociopragmatic strategies of
Native Speakers of Malay (NSM) Malay L1 English L2
Alerter 36 35
Illocutionary force indicating device 52 51
Taking/minimising responsibility 52 52
Explanation 1 2
Offer of repair 27 29
Promise not to repeat offence 1 2
Distracting strategy 12 14
*Closing statement 1 7
Thanking the receiver 4 4
*Expressing concern/empathy 10 2
Note: *Fisher’s exact test (one-tailed) shows a significant difference at p < 0.05.
Open Access
by Fisher’s exact tests, p < 0.05. When each group was ana-
lysed separately, the Native Speakers of Malay used a signifi-
cant number of Low-Risk but no High-Risk Distracting strate-
gies, both in L1 Malay (HR = 0, LR = 12) and in L2 English
(HR = 0, LR = 14). The Native Speakers of English used sig-
nificantly more High Risk than Low Risk Closing statements
(HR= 15, LR = 5). When L1 performance was compared, Na-
tive Speakers of Malay used significantly more L1 Low Risk
Distracting strategies than their NSE counterparts (NSM = 12,
NSE = 5). Unlike the Native Speakers of Malay in L1, the Na-
tive Speakers of English used a significant number of Closing
statements in L1 High Risk situations (NSM = 0, NSE = 15)
(but the low number of Low Risk Closing statements did not
differ significantly in the two groups’ L1 use (NSM = 1, NSE =
Pragmalinguistic Realisations of Strategies
This section examines the pragmalinguistic realisations of
those socio pragmatic categories where significant differences
between groups, first and second language or levels of risk
were found; it also examine show pragmalinguistic realisations
of other categories may contribute to answering the research
Closing Statements
In the two L1s, the only significant difference between the
NSM and NSE groups was in the English native speakers’ fre-
quent use of closing statements. In addition, there is evidence
that the NSEs and the NSMs were using this strategy in very
different ways. Most of the English L1 Closing statements were
used in High Risk situations. These statements were either
conventional complimentary closes like “Best wishes” and
“Regards” (in 5 of the 15 NSE responses) or a quasi-signature,
i.e. the name of the sender (in all 15 of the responses), occur-
ring in all cases at the end of the message. This clearly indicates
that some NSEs in the study conceptualise high risk text mes-
sage apologies as written medium texts requiring adherence to
formal conventions. On the other hand, in only 6 of the 15 HR
responses where NSEs made these statements did they begin
the message with a formal Alerter, such as a title or “Dear X”
or both; their writers may have been responding to the enforced
brevity of the text message, or they may have been, consciously
or not, crafting a message that was a hybrid between writing
and speech. The hybridity of this text creation is reinforced by
the fact that these same messages also contain approximations
of speech such as ellipsis (“Just wanted to say again how sorry I
am”) and indications of prosody (“I’m *very* sorry”).
How did this use of Closing statements by the NSEs compare
to the NSMs’ use of the strategy? Only one NSM used a Clos-
ing statement in L1, in a Low Risk situation, and this was a
code-shift to English (“Luv ya. X.”). NSMs used quasi-signa-
tures three times in L2 English, all in HR situations; but they all
came at the beginning of the message (“Hello mdm. this is
reza”; “Dr, its Sara”; “Hi, Mrs Smith, this is maya”). The
NSMs’ use of quasi-signatures in English is very different from
the NSEs’ use of them: it resembles phone conversation con-
ventions. Likewise, although the four Closing statements that
NSMs used in LR situations were complimentary closes, they
were not formal ones: there were three tokens of “Cheers!” and
one of “Enjoy your weekend!”. Again, these are arguably more
like spoken than written language. In other words, not only did
the NSMs hardly use Closing statements in L1, but all of their
uses of this strategy were more speech-like than writing-like,
marking an even clearer difference between the two groups and
a more uniform use of this strategy by each group than the ini-
tial quantitative analysis indicates.
While there were no overall differences in the use of this
strategy, how it was realised did differ between the two groups.
Notably, the NSM used formal address in 23 cases in L1 Malay,
compared to only 4 cases of formal address by the NSE group—
all of these in HR situations, unsurprisingly. This seemed to
carry over to the NSMs’ L2 responses, with 12 cases of formal
address in English. Given the great importance in Malay culture
of being refined, and of appearing to be so, it is perhaps not
surprising that the Malay respondents used formal address even
in their English texts. However, formal forms of address co-
existed with abbreviations and speech-like language in the same
texts, underlining once again the hybrid nature of these texts.
Minimising Responsibility: Avoiding Mentioning
the Offence
Although more often than not the offence was explicitly
mentioned as a means of acknowledging responsibility, par-
ticipants in both groups also frequently avoided mentioning the
offence too directly (e.g. “I thought that our meeting will be
tomorrow and I just realized it just now”; “[I’m sorry for] not
been able to meet you today”. “i js realized tht I hvnt returned
ur book 2libry yet”. There was no indication of a difference
between the two groups in this area. This less explicit way of
alluding to the offence might be seen as appealing to the kind of
common ground that is more usual in spoken than in written
exchanges. Interestingly, both groups pushed their responsibility
into the background in this way more often in High Risk than
Low Risk situations (NSE: HR = 12, LR =4; NSM L1: HR = 10,
LR =3; NSM L2: HR = 15, LR = 7). This may reflect the greater
threat to the apologiser’s face posed by a High Risk apology.
Expressing Concern or Empathy
One difference in L1 and L2 use by the NSMs was in the
number of expressions of concern/empathy (10 in L1, of which
4 in HR, 6 in LR situations; only 2 in L2, both in LR situations).
In L1 HR situations the concern regards the receiver’s having
had to wait unnecessarily for the apologiser. In the L1 LR
situations, concern is also expressed about making the receiver
wait (twice), but also about disturbing the receiver (twice),
about the receiver’s whereabouts (once) and generally about the
receiver’s well-being (once). This increased attention to the
receiver’s feelings in the L1 Malay situations over the L2 situa-
tions appears to correspond to Goddard’s (1997) observation
that in Malay culture it is important to be considerate towards
other people’s feelings without having to be alerted to these:
“Part and parcel of being brought up Malay is learning to an-
ticipate others” wishes and, as far as possible, to accommodate
them (Goddard, 1997: p. 194). The increased attention to other
people’s feelings is what would be predicted if texting in Malay
evoked Malay cultural values for the NSMs in a way that tex-
ting in English did not.
Distracting Strategies
The L1 Malay speakers used a significant number of low-risk
Open Access 301
but no high-risk Distracting strategies, in L1 Malay (LR = 12,
HR = 0) and in L2 English (LR = 14, HR = 0).
(The L1 English speakers only used 6 Distracting moves, but
the difference between these and the L1 Malay users’ Distract-
ing moves did not reach significance.) Most of the Distracting
moves (9 out of 12 for L1 Malay, 10 out of 14 for L2 English)
were requests to the offended party for something other than
forgiveness, e.g. “Plz,plz say u’r not mad at me :p”, “Call me
back please?”, “Erm, pls ignore it”; there were also some in-
stances of humour and of offers of unrelated gifts or favours (“I
will buy you dinner”, “aku belanja makan”). The Malay users’
extensive use of distraction strategies relates well to Malay
cultural scripts as described by Goddard (1997). Firstly, all of
these ways of distracting allow the apologiser to avoid making
direct reference to the offence, which is highly preferable in
Malay culture to avoid loss of face for the apologiser. Secondly,
given the Malaysian social stricture against expressing wishes
about what other people should do, it may well be the perceived
higher social status of the receivers in the HR situations that
prevent NSM participants from asking them to say or to do
something and means that all of the NSM distracting moves
were in Low Risk situations. It is interesting that the Malay
users have not differentiated between the L1 and L2 situations
here, given that they have done so for some other pragmalin-
guistic realisations; perhaps hierarchical relationships tend to
feel similar between their two cultures.
Written or Spoken, or Hybrid?
This study has provided evidence that text apologies can in-
deed be characterised as a hybrid genre; and differentially so
for different communities of users. One indication that texts
resemble spoken discourse is the prevalence of moves in the
corpus where members of both groups avoided mentioning the
offence they were texting about when communicating in L1,
contrary to the expectation of explicitness in written text. Omit-
ting to mention the offence was more frequent in high-risk than
in low-risk situations, and whereas in other cases in the study,
high-risk situations seemed to prompt more formal, more writ-
ten-like text, in the domain of mentioning the offence this was
not the case. It is striking that both of the language/culture
groups avoided mentioning the offence in their L1 apologies. It
will be interesting to investigate text apologies in other L1s, to
examine whether this lack of explicitness in referring to an
offence is a common aspect of this aspect of the developing
genre across cultures.
An even more powerful demonstration of the hybrid nature
of these texts is the case of the openings and closings of the
high risk text apologies. When finishing their high risk apology
texts, the English L1 users employed complimentary closes that
would have been appropriate in a traditional letter, and often
added quasi-signatures as well; yet these same writers were far
from systematic in beginning these texts formally. In contrast,
the Malay L1 users never used formal complimentary closes,
and when they did give their names in their texts, they did so at
the beginning, as if in a telephone call; at the same time, they
used formal letter-type salutations and professional titles, often
in their first language texts and sometimes in those they wrote
in the second language. These two groups of texters appear to
have developed distinctive language/culture-specific genre
conventions which, in different ways for each culture, posi-
tioned the text apology somewhere between writing and speech.
Culturally Specific Pragmatics
In this study’s examination of the differences between the
first-language sociopragmatics of the two groups and the ways
in which strategies were realised pragmalinguistically, an initial
impression of uniformity gave way to a more nuanced picture.
For example, on a macro level, there appear to be few differ-
ences between the L1 text messages of native speakers of Ma-
lay and the L1 text messages of native speakers of English.
Both groups use the same sociopragmatic strategies, and for all
but one strategy (Closing statements) the frequencies of use
were not significantly different. However, closer examination
reveals that the pragmalinguistic realisations of these strategies
did differ, sometimes in important ways. For example, while
the difference did not reach significance, NSEs did not express
concern or empathy as often as NSMs; and more importantly,
the NSMs expressed concern or sympathy significantly more
often in L1 Malay than they did in L2 English, in ways that
corresponded well to the cultural scripts that Goddard (1996,
1997) has articulated to characterise Malay culture. This sug-
gests that the Malay native speakers were, consciously or not,
tailoring their responses to the perceived norms of two different
In the case of strategies for distracting from the offence, a
different pattern obtained. Again, the NSEs used fewer of these
strategies than the NSMs, and again without this difference
reaching significance. However, what is interesting here is that
while the NSMs made a clear distinction between low risk and
high risk situations—with abundant use of these strategies in
low risk situations and none in high risk ones—they did this in
the same way in first and second languages. Although they are
highly proficient speakers of English, living and working in an
English speaking environment, it appears that the NSMs retain
in the second culture those first-culture norms which prompt
them to distract interlocutors with a request for action in a low
risk situation, but not in a high-risk one, when the offence is a
serious one with substantial face-threatening potential for the
interlocutor. This runs counter to the conclusions of Maros
(2006), who argued that adherence to first-culture norms in
apologies was probably an effect of her participants having
spoken English almost exclusively with Malay interlocutors.
The current study suggests that some cultural norms may be
more resistant to acculturation than others.
These two examples of expressions of concern/empathy, on
the one hand, and distracting strategies, on the other, reveal a
nuanced negotiation of second language/second culture mem-
bership by L2 users navigating their two worlds in a complex
way, sometimes accommodating to the second culture and
sometimes retaining first culture pragmatic usages. It is only
detailed study of particular cases that allows this kind of insight
to emerge.
Implications and Future Research
The phenomena that have come to light in the present study
have obvious implications for cross-cultural texting. In a cross-
cultural text apology, differing practices and expectations have
the potential to lead to the offended party’s perceiving the be-
haviour of the apologiser as inappropriate, with potentially
Open Access
serious consequences. The resistance of the acculturated L2
English texters in the current study to alteration of some aspects
of their first-culture apology convention suggests that implicit
learning of new norms in this area may not be effective. There-
fore, an obvious next step is to examine perceptions of cross-
cultural texters receiving apologies that do not correspond to
their (possibly unconscious) expectations. This in turn can lead
to the incorporation in language teaching materials of aware-
ness raising with regard to these issues. In a situation where
English is a lingua franca, it may be worth carrying out similar
studies with each of the first language/culture groups involved,
to inform the development of appropriate teaching materials.
Afghari, A. (2007). A sociopragmatic study of apology speech act
realization patterns in Persian. Speech Communication, 49, 177-185.
Aijmer, K. (1996). Conversational routines in English. Convention and
creativity. New York: Addison Wesley Longman Limited.
Ancarno, C. (2005). The style of academic e-mails and conventional
letters: Contrastive analysis of four conversational routines. Iberica,
9, 103-122.
Badrul Redzuan, A. H. (2006). New media language: “Txtgda trickster
talk”. In L. S. Chin, & T. K. Hua (Eds.), Composing meanings: Me-
dia text and language (pp. 41-59). Bangi: UKM.
Baron, N. S. (2000). Alphabet to email: How written English evolved
and where it’s heading. London: Routledge.
Baron, N. S. (2004). Rethinking written culture. Language Sciences, 26,
Bataineh, R. F., & Bataineh, R. F. (2008). A cross-cultural comparison
of apologies by native speakers of American English and Jordanian
Arabic. Journal of Pragmatics, 40, 792-821.
Bharuthram, S. (2003). Politeness phenomena in the Hindu sector of the
South African Indian English speaking community. Journal of Prag-
matics, 35, 10-11.
Blum-Kulka, S., House, J., & Kasper, G. (1989). Cross-cultural prag-
matics: Requests and apologies. Norwood, NJ: Ablex Publishing
Bohnke, C. (2001). On the acquisition and learning of linguistic-prag-
matic competence exemplified by the speech act “apologizing”.
Moderna Sprak, 95, 165-191.
Brown, P., & Levinson, S. (1978). Universals in language use: Polite-
ness phenomena. In E. N. Goody (Ed.), Questions and politeness (pp.
56-289). Cambridge: Cambridge University Press.
Carter, R., & McCarthy, M. (2006). Cambridge grammar of English.
Cambridge: Cambridge University Press.
Chang, W. (2008). Australian and Chinese perceptions of (im)polite-
ness in an intercultural apology. Griffith Working Papers in Prag-
matics and Intercultural Communication, 1, 59-74.
Clyne, M., Fernandez, S., & Muhr, R. (2003). Communicative styles in
a contact situation: Two German national varieties in a third country.
Journal of German Linguistics, 15, 95-154.
Cohen, A. D., & Olshtain, E. (1981). Developing a measure of so-
ciocultural competence: The case of apology. Language Learning, 31,
Cohen, A. D., & Shively, R. L. (2007). Acquisition of requests and
apologies in Spanish and French: Impact of study abroad and strat-
egy-building intervention. Modern Language Journal, 91, 189-212.
Collot, M., & Belmore, N. (1996). Electronic language: A new variety
of English. In S. C. Herring (Ed.), Computer mediated communica-
tion: Linguistic, social and cross-cultural perspective (pp. 13-28).
Philadelphia, PA: John Benjamins.
Crystal, D. (2005). How la n gu a g e w orks. London: Penguin Books.
Crystal, D. (2008a). Texting. ELT Journal, 62, 77-83.
Crystal, D. (2008b). Txtng. The gr8 db8. Oxford: Oxford University
Danet, B., & Herring, S. C. (2007). The multilingual internet: Lan-
guage, culture, and communication online. Oxford: Oxford Univer-
sity Press.
Ely, R., & Gleason, J. B. (2006). I’m sorry I said that: Apologies in
young children’s discourse. Journal of Chi ld L angu age, 33, 599-620.
Gains, J. (1999). Electronic mail—A new style of communication or
just a new medium?: An investigation into the text features of e-mail.
English for Specific Purposes, 18, 81-101.
Gimenez, J. C. (2000). Business e-mail communication: Some emerg-
ing tendencies in register. English for Specific Purposes, 19, 237-251.
Goddard, C. (1996). The “social emotions” of Malay (BahasaMelayu).
Ethos, 24, 426-464.
Goddard, C. (1997). Cultural values and “cultural scripts” of Malay
(BahasaMelayu). Journal of P r agm atics, 27, 183-201.
Grieve, A. (2010). “Aberganzehrlich”: Differences in episodic structure,
apologies, and truth-orientation in German and Australian workplace
telephone discourse. Journal o f Pragmatics, 42, 190-219.
Hatipoğlu, Ç. (2004). Do apologies in e-mails follow spoken or written
norms? Some examples from British English. Studies about Lan-
guages, 5, 21-29.
Hobbs, P. (2003). The medium is the message: Politeness strategies in
men’s and women’s voice mail messages. Journal of Pragmatics, 35,
Holmes, J. (1989). Sex differences and apologies: One aspect of com-
municative competence. Applied Linguistics, 10, 194-213.
Intachakra, S. (2004). Contrastive pragmatics and language teaching:
apologies and thanks in English and Thai. RELC Journal, 35, 37-62.
Kampf, Z., & Blum-Kulka, S. (2007). Do children apologize to each
other? Apology events in young Israeli peer discourse. Journal of
Politeness Research, 3, 11-37.
Kasanga, L. A., & Lwanga-Lumu, J.-C. (2007). Cross-cultural linguis-
tic realization of politeness: A study of apologies in English and
Setswana. Journal of Politeness Research, 3, 65-92.
Kim, H. (2008). The semantic and pragmatic analysis of South Korean
and Australian English apologetic speech acts. Journal of Pragmatics
40, 257-278.
Leech, G. (1983). Principles of pragmatics. London: Longman.
Ling, R., & Baron, N. S. (2007). Text messaging and IM: linguistic
comparison of American College Data. Journal of Language and So-
cial Psychology, 26, 291-298.
Linnell, J., Porter, F. L., Stone, H., & Chen, W. (1992). Can you apolo-
gize me? An investigation of speech act performance among
non-native speakers of English. Educational Linguistics (University
of Pennsylvania Graduate School of Education), 8, 33-53.
Maros, M. (2006). Apologies in English by adult Malay speakers: Pat-
terns and competence. The International Journal of Language, Soci-
ety and Culture, 19, 1-14.
Mey, J. L. (2001). Pragmatics: An introduction. Oxford: Blackwell.
Mulamba, K. (2009). Social beliefs for the realization of the speech acts
of apology and complaint as defined in Ciluba, French and English.
Pragmatics, 19, 543-564.
Nakano, M., Miyasaka, N., & Yamazaki, T. (2000). A study of EFL
discourse using corpora: an analysis of discourse completion tasks.
Journal of Pan-Pacific Association of Applied Linguistics, 4, 273-
Nureddeen, F. A. (2008). Cross cultural pragmatics: Apology strategies
in Sudanese Arabic. Journal of Pra g m at i cs , 40, 279-306.
Open Access 303
Olshtain, E. (1989). Apologies across languages. In S. Blum-Kulka, J.
House & G. Kasper (Eds.), Cross-cultural pragmatics: Requests and
apologies (pp.155-173). New Jersey: Ablex Publishing Corporation.
Olshtain, E., & Cohen, A. (1983). Apology: A speech act set. In N.
Wolfson, & E. Judd (Eds.), Sociolinguistics and language acquisi-
tion (pp. 18-36). Rowley, MA: Newbury House.
Park, H. S., Lee, H. E., & Song, J. A. (2005). “I am sorry to send you
SPAM”. Cross-cultural differences in use of apologies in email ad-
vertising in Korea and the U.S. Human Communication Research, 31,
Rose, K. R. (2000). An exploratory cross-sectional study of interlan-
guage pragmatic development. Studies in Second Language Acquisi-
tion, 22, 27-67.
Sabate i Dalmau, M., & Curell i Gotor, H. (2007). From “Sorry very
much” to “I’m ever so sorry”: Acquisitional patterns in L2 apologies
by Catalan learners of English. Intercultural Pragm a ti c s, 4, 287-315.
Scher, S. J., & Darley, J. M. (1996). How effective are the things peo-
ple say to apologize? Effects of the realization of the apology speech
act. Journal of Psycholinguistic R esearch, 26, 127-140.
Shardakova, M. (2005). Intercultural pragmatics in the speech of
American L2 learners of Russian: Apologies offered by Americans in
Russian. Intercultural Pragmatics, 2, 423-451.
Shively, R. L., & Cohen, A. D. (2008). Development of apologies and
requests during study abroad. Ikala, Revista de Lenguaje y Cultura,
13, 57-118.
Sugimoto, N. (1997). A Japan-U.S. comparison of apology styles.
Communication Research, 24, 349-369.
Supyan, H. (2006). The features in Malay SMS Texts. In L. S. Chin, &
T. K. Hua (Eds.), Composing meanings: Media text and language
(pp. 61-73). Bangi: UKM.
Trosborg, A. (1995). Interlanguage pragmatics. Requests, complaints
and apologies. Berlin: Mouton de Gruyter.
Warga, M., & Scholmberger, U. (2007). The acquisition of French
apologetic behavior in a study abroad context. Intercultural Prag-
matics, 4, 221-251.
Wierzbicka, A. (1991). Cross-cultural pragmatics: The semantics of
human interaction. New York: Mouton de Gruyter.
Wierzbicka, A. (1996). Semantics: Primes and universals. Oxford:
Oxford University Press.
Woodfield, H. (2008). Problematising discourse completion tasks:
Voices from verbal report. Evaluation and Research in Education, 21,
Wouk, F. (2005). The language of apologizing in Lombok, Indonesia.
Journal of Pragmatics, 38, 1457-1486.
Yang, T.-K. (2002). A study of Korean EFL learners’ apology speech
acts: Strategy and pragmatic transfer influenced by sociolinguistic
variations. Journal of Pan-Pacific Association of Applied Linguistics,
6, 225-243.
Zhang, H. (2001). Culture and apology: The Hainan Island incident.
World Englishes, 20, 383-391.
Open Access
Appendix A
Texts Sent to Participants
HR1: U were supposed to meet ur lecturer (female) at 2pm.
However, u only remembered at 5pm. U called 2 apologise bt
she was not in. Send an apology thru SMS.
HR2: Ur supervisor (female) asked u to return a library book
which she borrowed. U agreed to help bt forgot to return d book.
A week later she sends a text msg/SMS inquiring whether the
book has been returned. Reply to the text msg to apologise.
LR3: U were supposed to meet a female cousin at 11am, but u
only remembered around 2pm. U called but it was not answered.
Apologise thru a text msg/SMS.
LR4: U send an SMS from mobile to a friend, but when d reply
comes back u realise tht u hv sent it to a stranger. Send an SMS
to d person to apologise.
HR1: Anda membuat temujanji dgn seorg pensyarah perem-
puan pd pkl 2 ptg, ttp hanya terigt pd pkl 5 ptg. Anda mnelefon
pensyarah itu utk meminta maaf ttp dia tiada d pejabat. SMS
pensyarah itu utk meminta maaf.
HR2: Pensyarah penyelia anda (perempuan) mminta anda me-
mulangkn buku ppustakaan yg dipinjam olehnya. Anda brsetuju
utk mbantu ttp lupa utk memulangkn buku tsebut. Sminggu
kmudian, dia mhantar SMS bertanyakn ttg buku itu. Bls SMS
tsb utk mminta maaf.
LR3: Anda bjanji dgn sepupu perempuan anda utk bjumpa pd
pkl 11 pg, ttp anda hanya teringat janji itu pd pukul 2 ptg.
Sepupu anda tidak mjawab pggilan telefon anda. Hantar SMS
kpd sepupu anda utk meminta maaf.
LR4: Anda mhantar mesej (SMS) kpd seorg rakan, ttp apabila
anda mdapat balasan SMS tsebut, anda mdapati yg anda ter- salah
hantar kpd org yg tidak dikenali. Hantar SMS utk mminta maaf.
Appendix B
Coding Scheme
Sociopragmatic strategies Pragmalinguistic realisations and examples
1) Alerter
1a) greetings (hi, hello, hey, good morning, salam)
1b) formal address (Dr., Professor, Madam)
1c) name
1d) formal phrasing (Dear Mr., Dear Dr.)
1e) endearments (babe, love)
2) Illocutionary Force Indicating Device (IFID) 2) I’m sorry, I apologise, Please forgive me, Maaf
3) Taking responsibility
(or attempts at minimising responsibility)
3a) explicit mention of offence (I forgot...Saya terlupa, Wrong number! ... salah hantar mesej)
3b) implicit mention of offence (It slipped my mind, ...tidak ingat, meant to send that to..., saya nak hantar kepada...)
3c) avoiding mentioning offence (I missed the meeting, …tidak menghadirkan diri...)
3d) explicit self-blame/reproach (My mistake, it was wrong of me ... kecuaian saya)
3e) earlier failed attempt at mitigating offence (I tried to call… I called but... Saya cuba call Puan)
3f) expression of embarrassment (I feel awful/bad)
3g) self-defence/validation (This is not me, I was busy with…saya sibuk dengan...)
3h) justify receiver’s reaction to offence (You must be mad at me... mesti marah kan)
3i) implicit self-blame/reproach (I should return it last week, sepatutnya saya buat minggu lepas...
4) Explanation/account 4) external, out-of-control circumstances which led to the offence (I was trapped in the elevator for 3 hours, my
sister was involved in an accident, adahaltadi, ada org datang...)
5) Offer of repair
5a) compensatory offer directly related to offence
(Could I arrange another appointment? Boleh buat temujanji lain?)
5b) additional compensatory offer directly related to offence, may explicitly take into account receiver’s con-
venience (If it’s possible..., to your convenience, I’ll pay for the fines, Sekiranya tak menyusahkan...).
6) Promise to not repeat offence 6) (This will not happen again, I will be there, saya janji akan datang)
7) Distracting from offence
7a using humour (I must’ve been blind, ... nanti kene jual...)
7b appeaser (I will buy you dinner, aku belanja makan)
7c plea/request to receiver for something other than forgiveness
(I hope you understand? Ignore the message, tolong abaikan)
8) Closing statement
(both formal and informal–wish-like)
8a) complimentary close (Regards, Best wishes, Enjoy your weekend, Take care)
8b) sender’s name/initials
9) Thanking receiver 9) (Thanks, tq, Terima kasih)
10) Concern for receiver/show of empathy 10) (I know you’re busy, are u ok? ... buat puan tertunggu2, lama tak tunggu?)
11) Others Intensifiers of apology Red
i) IFID internal intensifier Bold (very, truly, extremely, so, sangat, banyak2)
ii) Emotional expression/exclamation Underlined (Whoops, OMG, Alamak)