Use and Standardisation of Mauritian Creole in Electronically Mediated Communication



New technologies have affected language use and attitudes in many communities. Kreol, a French-lexified Creole and the nonstandardised first language of the majority of Mauritians, is now gaining ground as a written language in the specific context of electronic-mediated communication. This has led to the emergence of writing norms among users of the language. These norms are founded on etymological phonemic and mixed conventions. This study based on data gathered through questionnaires analyses the attitudes of 66 young Mauritians towards the three orthographies used in electronically mediated communication and the standardisation of the language in new technologies. It also briefly discusses some of the spelling conventions used in Internet postings. I show that the etymological system is perceived as most readable, learnable and closest to French. Users believe that Kreol can act as a unifying factor among different ethnolinguistic groups in Mauritius. Responses also highlight the potential of users and electronic-mediated communication in bringing about the standardisation of Kreol.


The advent of new means of communication like text messages, e-mails, and Internet chats has led to sociocultural, economic, and linguistic changes in many communities (e.g., Crede, 1995; Choi & Danowski, 2002; Nishimura, 2004; Ho, 2005). Mauritius, a multilingual and multiethnic nation in the Indian Ocean, has also experienced changes at the sociocultural, political, economic and linguistic levels resulting from the use of new technologies. However, there is limited research on the impact of electronically mediated communication on the national linguistic landscape (Rajah-Carrim 2004).

This paper makes a contribution in this area by looking specifically at the effect of new technologies on Mauritian Creole, locally known as Kreol. It is based on a small-scale pilot study which looked at young Mauritians' perception of writing systems used in computer-mediated communication (CMC). Data was gathered through questionnaires. We analyse the way in which CMC affects use of and attitudes to this nonstandard language. Previous research has looked at the way in which the written norms of standard languages are affected by new technologies (Collot & Belmore 1996, Baron 2000). This paper addresses the issue of standardisation from a different perspective in that it assesses whether the increasing use of Kreol in CMC is leading to the emergence of written norms for the language.

This empirical study also analyses whether there is any relationship between the various orthographies' readability, learnability, and closeness to French. This study is particularly significant given that Kreol has not been officially standardised yet. It will be useful if speakers' views on Kreol are taken into consideration when devising an orthography for the language.

This paper is structured as follows: The following section places this study within a sociolinguistic background. The third section focuses on the study: methods and results. The fourth section discusses some examples of the conventions used for writing Kreol in new technologies. In the fifth section, I discuss the data and relate it to issues raised in Section 2. The fifth section acts as a conclusion.


Mauritius is a nation of immigrants, where loyalties usually lie with specific ethnolinguistic groups rather than the nation as a whole. There is a growing awareness that Mauritians need symbols of a shared identity in order to develop a sense of nationhood. Language is such a symbol (Eriksen 1998, Suleiman 2003, Joseph 2004). A shared language can create a sense of identity and cohesion among people of different ethnolinguistic and religious backgrounds. Kreol could function as the official national language of Mauritius in that it is the one language that is widely spoken on the island by people of all ethnic and religious groups. It therefore transcends ethnic barriers and unites people from various ethnolinguistic communities.

Kreol is the first language of at least 75% of the population of Mauritius (Central Statistical Office, 2000). It exists alongside European languages like English and French and ancestral languages like Bhojpuri, Hindi, Marathi, Mandarin, and Urdu. These languages do not function only as media of communication but also markers of identity. In fact, many of the ancestral languages are hardly spoken on the island and serve almost exclusively as markers of ethnic and/or religious identity (Hookoomsing 2003).

However, there are obstacles to the promotion of Kreol as a national language. First, the language has an ethnic index in that it is sometimes associated with the Afro-Mauritians, the descendants of slaves – locally known as Creoles (Carpooran 2004). Second, like many creoles, Kreol is seen as a broken language that cannot be used in formal domains and written interactions (Rickford & Traugott 1985). It tends to be compared to its lexifier French, seen as a legitimate language which can be used in a range of domains and has a fixed spelling system (Rajah-Carrim 2004). The absence of an officially approved standard writing system for the language is often quoted as a reason against its use in formal domains.

Some linguists and private institutions have devised standard forms for the language. The group Ledikasyon Pu Travayer (LPT) runs literacy classes in Kreol. They use their own dictionary which is based on phonemic principles, i.e., on the sounds that occur in the language without reference to the lexifier (Ah Vee, member of LPT, personal communication). The Church with the help of the linguist Dev Virahsawmy have also come up with an orthography for Kreol known as grafi legliz. Grafi legliz could be considered as an “intermediate phonemic orthography” (Siegel 2005: 147): It is largely based on phonemic principles but does make some concessions to French spellings (Virahsawmy, personal communication; also Hookoomsing 2004). For instance, in LPT orthography “drink” (French <boire>) is written as <bwar> but in grafi legliz as <boir>, “corner” (French <coin>) is <kwin> in LPT and <koin> in grafi legliz, “cook” (French <cuire>) is <kwi> in LPT and <koui> in grafi legliz. The latter orthography tends to avoid the use of <w> within syllables. Instead it opts for a combination of other graphemes, e.g., <oi> and <ou> above. The use of these vowels brings grafi legliz closer to French. However, both orthographies promoted by different institutions are phonemic rather than etymological. Neither of these spelling systems has official backing.

In 2004, the government set up a committee to devise a standard orthography, grafi larmoni, for the language. The committee submitted their report later that year. Hookoomsing (2004: 28) argues that Kreol

has now reached a remarkably high degree of convergence and stability in terms of orthographic standardization. The time has come for the elaboration of a harmonized orthography based on a palette of symbols currently in use and on the expectations of potential users, particularly in the education system.

The system proposed by Hookoomsing and colleagues draws on the orthographies currently in use and is largely phonemic in nature. After the report was submitted, users were asked to send in their comments and suggestions on grafi larmoni to the Ministry of Education. The involvement of users was a laudable initiative as they are the ones who have to accept and use the system. A few months after the submission of the report, there were national elections and a change in government. To date there has not been any progress and no further effort from the new government to promote an official standard for the language (Hookoomsing, personal communication). The standardisation of Kreol is an ongoing process and has produced some tangible results in the media and advertising companies. However, since grafi larmoni is not used generally by the wider public, we will assume that there is no official orthography for Kreol that is endorsed by the nation. The language is still largely perceived as a nonstandard language by its speakers (Rajah-Carrim 2004). It should be noted that the lack of an officially agreed orthography for Kreol has not deterred Mauritians from expressing themselves in print using various orthographies for the language: Numerous books and booklets have been published in Kreol.

Standardisation is also a sociopolitical exercise and rarely purely a linguistic one. Deciding which variety of a language to standardise and which orthography to use for the language is tied to ideologies of power and identity (Joseph 1987). Choosing one variety of a language over others to become the standard adds to the prestige of that variety and sets it apart from other varieties of the language. It becomes the proper or correct form of the language. Standardisation adds to the legitimacy and status of a language. Although it affects both the spoken and written forms of a language, it is more marked in the latter case.

The standardisation of creoles is fraught with issues of legitimacy, power and identity (Schieffelin & Doucet, 1998; Sebba, 2000; Romaine, 2005). In the case of these languages, a distance has to be established between the creole and its lexifier and a balance has to be struck between ideologies of identity and practicalities. The creole orthography also has to be easily accessible to the majority of its speakers (Siegel, 2005). Linguists working on the standardisation of creoles have two options: they can choose a phonemic or an etymological orthography. A phonemic orthography, as the LPT orthography for Kreol, is based on the sounds that occur in the creole. Minimal reference is made to the lexifier. An etymological orthography is based on the spelling conventions of the lexifier. There are different ideologies attached to each system (Siegel, 2005). If the creole is used alongside its lexifier in the community, establishing the creole as a language in its own right usually means distancing it from the lexifier to a maximum. In fact, one “of the particular problems most creoles are facing is the question of Abstand or autonomy with regard to their lexifier” (Mühleisen, 2005: 7). A system based on phonemic principles allows users to create a distance between the creole and its lexifier, thus reinforcing the creole's position as a legitimate language independent of its lexifier.

However, establishing this distance can initially mean a compromise in terms of readability. Users who have learnt how to write the lexifier can find it hard to read a creole based on phonemic principles diverging from those of the lexifier (Hookoomsing, 2004). As in the case of Kreol, users may already have some norms that they adhere to locally. Readability of a new orthography, therefore, “becomes a decisive factor for its acceptance by an audience who may be used to already established de-facto norms and traditions of spelling creole” (Mühleisen, 2005: 8). There is sometimes a gap between linguists' standardisation of a language and speakers' use of their language in written domains.

Interestingly, although Kreol has no officially recognised orthography, it is increasingly being used in emails, Internet chats and text messaging by laypeople (Ramkhelawon, 2005). The spelling conventions used in these forms of communication can range from ones that closely resemble French to those that are totally different from the lexifier (Rajah-Carrim, 2004). For instance, “I speak Creole to you” could be (1) mo cause créole avec toi, (2) mo koz Kreol avek twa or (3) mo cose creol avek toi. These examples show varying degrees of departure from their French equivalents: causer (talk/chat), créole (creole), avec (with), toi (you). In the first example, etymological conventions are used and the words are spelt in exactly the same way as they would in French. In the second example, a phonemic system is used. The commonly used graphemes <k,w,z> index Mauritian identity by obscuring the French origins of words. The third example uses mixed conventions combining principles from the etymological and phonemic systems.

Electronically mediated forms of communication, therefore, provide a space for users of nonstandard or lesser used languages to write their language and at the same time develop their own spelling norms. What is more in the case of Kreol is that these forms of communication are used by Mauritians –especially young Mauritians –of all ethnicities. That is, they create connections among young people of different ethnic groups through Kreol. Because CMC requires the use of a shared code, it promotes Kreol among Mauritians of various origins. The need to communicate effectively in new technologies leads to the backgrounding of ethnolinguistic identities in favour of a national identity. That is, when interacting with Mauritians of other ethnicities, young Mauritians have to resort to a common code rather than use their ethnic language. Together, therefore, Kreol and CMC unite Mauritians of different ethnolinguistic backgrounds.

Sociolinguistic studies of technologically mediated communication are fairly restricted (for a critique and review of sociolinguistic studies of CMC see Androutsopoulos, 2006a). Much of the research on language use and new technologies has focused on the way in which new technologies affect and blur the distinction between written and oral communication (Baron, 1998), the interaction between language used in CMC and such social categories as gender, ethnicity and social class (Herring & Paolillo, 2006; Androutsopoulos, 2006b) and the role of CMC in the preservation of endangered and lesser-used languages (Cook, 2004; Eisenlohr, 2004).

There is limited data on the interaction between CMC and language standardisation for non-standard languages. Romani, for instance, a nonstandardised language spoken by the Roma people across Europe, is increasingly used in new technologies, thus leading to some convergence in spelling conventions (Matras, 2005: 37). Because there are no standard spelling norms for Romani, writers can “experiment with terms and writing conventions” (ibid) while at the same time accommodating to their interlocutors. CMC, therefore, provide opportunities for writing nonstandard languages. Deuber & Hinrichs (2007: 23) specifically argue that “developments in CMC are broadening the scope for written use” of Creoles and pidgins, languages often restricted to the oral domain.

Hinrichs (2004 & 2006) has looked at the way in which CMC can lead to the emergence of orthographic conventions for Jamaican Creole (JC). Like Kreol, JC functions mainly as a spoken language limited to informal domains but is valued as a symbol of identity. According to Mair (2002: 31), JC is a “symbol of powerlessness and degeneracy, on the one hand, and symbol of solidarity, truth, and connection to the Afro-Creole folk tradition, on the other” (these attitudes are common to other creoles (see Rickford & Traugott, 1985)). Although linguists have devised standard orthographies for the language largely based on phonemic principles, there is no officially agreed standard. Journalists and writers make up their own English-based spelling conventions for the language (Mair, 2002). Hinrichs (2004 & 2006) observes that JC is also increasingly used by laypeople in the specific CMC context. The need to communicate effectively in the various forms of CMC is leading to the emergence of orthographic conventions among users of JC without any centralised language planning. He notes that the “emerging Creole orthography in CMC is based on [Standard English] orthography” and also a desire of users to assert their identity as Jamaicans (Hinrichs 2004: 101). There are interesting parallels between the standardisation of JC and that of Kreol. Both appear to be user-driven and indexed as markers of a local identity.

In this paper, we explore the relationship between the standardisation of Kreol and CMC and also look at users' attitudes towards various spelling conventions used. The conventions used can broadly be classified into three categories: phonemic, etymological, and mixed. As mentioned above, these conventions differ in terms of their closeness to French. We would expect different degrees of readability and learnability to be associated with the various conventions. We find the three conventions used on the Internet (e.g., forums on, chats on

In the following sections, we discuss the interaction between Kreol, technology, and identity from users' perspectives based on responses to questionnaires administered to college students. We focus on users' perception of language use. Given that Kreol is increasingly being used in new electronic domains, can CMC promote the standardisation of the language? Can standardisation be brought about by people at the grassroot level?

The Study


This pilot study is based on questionnaires completed by 66 Mauritians aged between 16 and 19 years attending Moris College2 situated in an urban area. The college, first coeducational school on the island, is one of 182 secondary schools in Mauritius (Ministry of Education and Scientific Research, 2005) and has over 1,000 students. It is among the top 25 schools of the country. Moris College was chosen because of its mixed student population: Males and females attending the school are from various ethnolinguistic, religious, and socioeconomic groups.

The breakdown of informants is given in Table 1.

Table 1.  Background of informants3
Variables Number of informantsPercentage of informants
 Other/Not specified28
 Other/Not specified46
Place of residenceRural1015
 Not specified46

The questionnaire

The questionnaire was divided into 3 parts. In part 1, informants were asked for their age, gender, place of residence, ethnicity, religion, use of text messages/e-mails /Internet chats, and language(s) used in CMC. In part 2, they had to rate six extracts with respect to their readability, learnability, and closeness to French. The extracts were taken from e-mails, discussion forums and the linguist Virahsawmy's homepage. Respondents were also asked whether the spelling system used in the given extract should be promoted. Part 3 of the questionnaire consisted of open-ended questions. To test reliability, the internal consistency of the questionnaire was measured using Cronbach's alpha coefficient. Alpha for the whole sample was equal to 0.877, showing acceptable results.

The three orthographies commonly used for Kreol – phonemic, etymological, and mixed – were presented in the extracts. Extract 1 served as an example and will not be discussed in this paper. Extracts 2 and 6 both use orthographies that are close to French. Extract 2, written by a 27-year-old female, comes from a corpus of personal e-mails:

Mo ena ene dernier assignment pou hand out (…) Mo classe fek fini la… it's nearly 6 p.m… ek p faire noir..l’hiver fini rentrer! Mo enan classe lot semaine. Embrasse to soeur pou moi…

I have to hand out a last assignment (… ) The classes are just over… it's nearly 6pm… and it's dark here… winter has arrived ! I have classes next week. Kiss your sister for me…

Extract 6 was posted on one of the forums on the website servihoo.

Premierema nous fine aprane croire ki ledikassion et so banne certifika li seul passeport a ene job, alors nous tous et surtou banne malere parents ape rinte le cor pou donne zotte zenfan ene certificat. 12 a 20 banane de sacrifices, sa bane parents la ena ene seule lespoir: ki so piti pou gagne ene job a travers ene diplom ou degre, malgre tous les preuves du contraire, cetadire ki chomeur ape augmenter de jour en jour.

First, we were made to believe that education and its certificates are the only passports to a job. So, all of us, especially the poor parents, are working hard to give their children a certificate. Twelve to 20 years of sacrifice. These parents have only one hope: that their child gets a job through a diploma or degree, although there are proofs otherwise, that is, that unemployment is increasing from day to day.

These extracts are largely etymological in nature. For instance, dernier, classe, faire, l’hiver, semaine, nous, certificat are all written as they would be in French. Based on the LPT dictionary, the “Creole” equivalents for these words would be dernie, fer, liver, semenn, nu, sertifika. Given the close correspondence between these words and French, we would expect Mauritians to find these extracts very readable.

Extracts 3 and 7 written by Dev Virahsawmy use the Church orthography (grafi legliz) promoted by the religious institution and himself.

Extract 3:

Pou tou dimoun dan vilaz li ti finn fini tas lor poto. Li ti fini gagn venntroi-zan e li ti ankor viv kot so mama. Pourtan depi laz sez-an li ti pe gagn demann me bien vit bann pretan ti pe fonn.

For everyone in the village, she was stuck. She was already 23 and she was still living with her mum. But since the age of 16, she was getting proposals but very quickly suitors disappeared.

Extract 7:

Sharam, gran lider enn nouvo roiyom pouisan ole de troi minit trankilite. Li bizin res tousel. Li finn gagn laviktoir me ena enn ti gou amer parla ki pe anpes li zouir so sikse foul-foul.

Sharam, great leader of a new and powerful kingdom, wanted a few minutes of quiet. He had to be on his own. He was victorious but there was something that was preventing him from fully enjoying his success.

This system is largely phonemic (Virahsawmy, personal communication) although it does have words that approximate their French spellings. For example, nouvo (new; French nouveau) according to the phonemic system is written as nuvo. The spelling used here is a compromise between the phonemic nuvo and the etymological nouveau. Nasalisation is represented by nn rather than ne like in French. The phonemic nature of these extracts could mean that it would be less reader-friendly than extract 2, for instance.

Extracts 4 and 5 use the phonemic system promoted by LPT.

Zedi dernye, LEDIKASYON PU TRAVAYER finn resevwar UNESCO LITERACY AWARD dan enn selebrasyon zwaye ki ti ena lye dan orl Minispalite Q.Bornes () Grup ABAEIM finn met enn zoli lanbyans avek lamizik, poem ek sante pandan seremoni la.

Last Thursday, LEDIKASYON PU TRAVAYER received the UNESCO LITERACY AWARD in a festive celebration at the Municipality of Q.Bornes (… ) The group ABAEIM added to the party with music, poems, and songs during the ceremony.

Dan kad kanpayn pu enn politik ekonomik alternativ, militan Lalit dan tu landrwa finn organiz kolaz lafis. Lafis la kontenir demand dan plan dirzans ki Lalit pe propoze fas a kriz ekonomik. Zot bann demand dirzans ki form parti program pu lekonomi alternativ Lalit. Lafis la dir: “Bizin Servi ‘Larzan Lerop’ pu KREE ANPLWA PA DETRIR ANPLWA”; “Trening plis ALOKASYON SOMAZ”; Deviz bizin retrun ar Labank Santral pu DIMINYE DEPRESYASYON Rs.”; Pu sak laburer V.R.S. LATER AGRIKOL LOR BAY”; “Zwenn MOBILIZASYON LALIT”.

In support of a campaign promoting an alternative economic policy, Lalit militants from all regions stuck posters. The posters call for urgent action proposed by Lalit to deal with the economic crisis. These demands form part of Lalit's programme for an alternative economy. The poster says : ≪ We have to use ‘Europe’s money’ to CREATE EMPLOYMENT NOT DESTROY EMPLOYMENT ≫; ≪ Training plus UNEMPLOYMENT BENEFITS ≫; Foreign currencies have to go back to the Central Bank to REDUCE DEPRECIATION of the rupee ≫; For all labourers AGRICULTURAL LAND ON LEASE ≫; ≪ Join the LALIT CAMPAIGN ≫.

The graphemes k,z, and w are widely used in these extracts. The orthographic conventions used here diverge from French. As such, we would expect such a system to be harder to read than the etymological one as Mauritians have not been taught how to read and write Kreol phonemically.

After each extract, informants were asked to rate the spelling system used. They were given the following options:

  • For readability –very easy to read, easy to read, difficult to read, very difficult to read

  • For learnability –very easy to learn, easy to learn, difficult to learn, very difficult to learn

  • For closeness to French –same as French, close to French, different from French

The readability criterion measures how accessible the extract is while the learnability criterion assesses how easy it would be for people to acquire the system used in the extract. Our focus here is on the spelling system: How easy is it for people to understand and learn? Readability scores, often used to assess the difficulty of textbooks used in schools, have not been exploited in studies looking at the standardisation of a language. Readability formulas can provide an objective assessment of the readability of a text. In the questionnaire, readability was measured subjectively in that it was assessed by respondents rather than formulae. Learnability and closeness to French were also measured subjectively.

The data

In this corpus, most informants make use of one form of CMC. Text messaging and e-mails are used by 95% and 68% of informants, respectively. Only 39% of informants chat on the Internet. Kreol is the most used language in text messaging, as shown in the table below.

Table 2 shows that many respondents use a combination of languages in CMC. Of the 26 respondents who chat on the Internet, most do so “weekly”. Chatting on the Net is therefore not a common activity for young Mauritians in this corpus. Unlike English and French, Kreol is not used on its own in Internet chats. This is interesting as the informal nature of Internet chats would be conducive to the use of Kreol, a language of informal social interactions.

Table 2.  Languages used in new technologies
LanguagesEmailsChatsText messages
Kreol only4 %013 %
English only33 %23 %10 %
French only13 %12 %3 %
Kreol & English11 %19 %16 %
Kreol & French0014 %
English & French15 %19 %2 %
Kreol, English & French15 %24 %38 %
Depends2 %4 %2 %
Not stated4 %4 %3 %

The use of Kreol only in e-mails is also limited. English is the most commonly used language in e-mail interactions. This is not exceptional: given that e-mails are less informal than chats, we would expect to find a more formal language. It is interesting to note that Kreol and French in combination are not used in either e-mails or chats. Further investigations are needed to explain this situation.

Of particular interest is the use of languages in text messages. Of the 63 informants who use text messages, 53 do so daily, five every other day, three weekly, and two “on an irregular basis”. On the basis of the data in Table 2, we could argue that text messaging is more likely to have an impact on linguistic practices than e-mails and chats because of its wider use. Kreol (alone or in combination with other languages) is the language most often used in text messages. This is likely to influence attitudes towards the language. Also, if young Mauritians use Kreol so frequently in text messaging, it is highly likely that over time they will devise or have already devised some shared norms for writing the language. Through the open-ended questions, we aim to find whether, as Hinrichs (2004: 83) puts it for JC, “[increased] written usage of the Creole is leading to orthographic standardization through conventionalization”.

Although our informants come from different ethnic backgrounds, most of them make use of at least one form of electronic-mediated communication. They are also familiar with the use of written Kreol. They are therefore well placed to answer the questions regarding the standardisation of the language and its use in CMC. In the section below, we assess informants' perception of the different orthographies in the six extracts presented above.

The extracts

Given that Mauritians are not taught how to read and write Kreol, we hypothesise that an etymological system would be perceived as more readable than a phonemic one. The similarities with French, a language learnt by all Mauritians, would facilitate the reading of a system based on etymological principles. By contrast, a phonemic system would look unfamiliar and hence harder to read. However, note that for people who cannot read the lexifier language, the phonemic system is easier to learn when acquiring literacy because of its consistency and because new readers tend to decode sound by sound. In contrast, the etymological orthography preserves the inconsistencies and historical forms unrelated to pronunciation that are found in the lexifier language (Siegel 2005: 147).

All our informants are literate in French. Therefore, it is more likely that they would find the phonemic system harder to read.

Informants were asked to rate the various extracts on three main scales. A value was assigned to each response. The following options were offered (the figure in between parentheses shows the arbitrary value assigned to the response):

  • Readability –Very easy to read (1), easy to read (2), difficult to read (3), very difficult to read (4)

  • Learnability –Very easy to learn (1), easy to learn (2), difficult to learn (3), very difficult to learn (4)

  • Closeness to French –Same as French (1), Close to French (2), Different from French (3)

  • They were also asked whether the orthography presented in the extract should be promoted. In the following paragraphs, I present and discuss the findings.

Extract 2 –Etymological system

Table 3 shows that most informants thought that this etymological orthography was very easy to read. However, they believed that such a system would be difficult to learn. This extract was rated as easy or very easy to read and difficult or very difficult to learn by the greatest number of informants. Also, although this system is largely etymological in nature, 78% of respondents said that it was different from French.

Table 3.  Values assigned to Extract 2
Characteristics1234No response
Closeness to French1%20%78%N/A1%

Extract 3 –Grafi Legliz (Phonemic)

Being a phonemic system, grafi legliz should be perceived as more difficult to read and learn than the orthography used in Extract 2. This is confirmed by our data: more informants rated Extract 3 as difficult to read and different from French than they did for Extract 2. Forty-four percent of informants thought that grafi legliz was very easy or easy to learn compared to 19% for the etymological spelling.

Extract 4 –Grafi LPT (Phonemic)

Unlike grafi legliz, the LPT orthography does not use any French conventions at all. Fifty-nine percent of informants found the phonemic orthography very easy to read and easy to read. This extract was considered as more difficult to read than extracts 2 and 3. However, compared to the two previous extracts, it was rated more highly in terms of learnability and as expected, difference from French. Similar observations apply to Extract 5 which uses the same phonemic orthography.

Extract 5 –Grafi LPT (Phonemic)

Although both extracts 4 and 5 use the same orthographic conventions, more informants found Extract 5 difficult and very difficult to read and learn than Extract 4. The difference in topic of the two extracts could explain the slightly different ratings: Extract 5 is more technical than Extract 4.

Extract 6 –Etymological spelling

Like Extract 2, Extract 6 uses conventions that closely resemble French. Table 6 shows that respondents believed that this extract was closer to French than Extract 2. Although the orthography used here shows more similarities to the French orthography than Extract 2, it is not considered as easy to read as Extract 2: 77% of informants thought that this extract was very easy or easy to read compared to 96% for Extract 2. However, more informants believed that this extract would be very easy or easy to learn. In fact, the figures show that this specific orthography is the one that is the closest to French and would also be the most easily learnt one.

Table 6.  Values assigned to Extract 5
Characteristics1234No response
Closeness to French011%80%N/A9%

Extract 7 –Grafi Legliz

This extract like Extract 3 uses a phonemic spelling system. Unlike Extract 3, this extract was considered to be more difficult to read. In fact, it was rated as the most difficult to read extract. Its figures for learnability are very close to those for Extract 5 (the LPT phonemic orthography). Like Extract 3, more than 50% of the informants believed that this extract would be difficult or very difficult to learn. Also, this extract was rated as different from French by the largest number of informants.

Comparing the different systems

Table 9 shows the values assigned for readability, learnability, and closeness to French for all extracts.

Table 9.  Values assigned to all extracts
CharacteristicsValues assignedExtract 2 (%)Extract 3 (%)Extract 4 (%)Extract 5 (%)Extract 6 (%)Extract 7 (%)
Closeness to French1101000


If we regroup the extracts together on the basis of the conventions used, we find that the phonemic system promoted by LPT and exemplified by extracts 4 and 5 has been rated as the most difficult to read. It is very closely followed by grafi legliz, another phonemic orthography. Could grafi legliz's partial accommodation to French orthography explain why this system is deemed more readable than the LPT one? We come back to this point when we discuss closeness to French. The extracts that show the closest resemblance to French, i.e., extracts 2 and 6, are perceived as the ones that are easiest to read.


The etymological system is deemed as the most learnable one. Seventy-two percent of informants thought that extracts 2 and 6 would be very easy or easy to learn compared to 43% and 39% for the extracts using the LPT orthography and grafi legliz respectively. Only 23% of respondents believed that the etymological system would be difficult to learn –compared to 56% for grafi legliz and 51% for LPT orthography. Once more, we see that the two phonemic systems tend to be perceived in the same way. Overall, however, more informants rated grafi legliz as difficult or very difficult to learn than LPT orthography.

Closeness to French

Informants thought that all extracts showed some differences from French. As would be expected, more informants thought that extracts 2 and 6 were close to French than extracts 3, 4, 5 and 7. Of the two phonemic systems, grafi legliz was perceived as more different from French than LPT orthography: 87% and 8% thought that grafi legliz was different from French and close to French, respectively, compared to 81% and 10% for LPT orthography. Therefore, although grafi legliz employs some French spelling conventions, users still perceive it as different from French. It is interesting that LPT orthography which does not use any French conventions is perceived as closer to French than grafi legliz.

Readability and learnability

The above analysis suggests that there is a relationship between readability and learnability. The systems that are perceived as more readable also tend to be perceived as more learnable. The results for readability and learnability were combined for each extract, as in Table 10.

Table 10.  Extract 7: Readability and learnability values

A Pearson's correlation coefficient test was performed on the data in Table 10 and similar data for the other extracts. The tests reveal that for all but Extract 2 there is a significant relationship between readability and learnability (Extract 3: Pearson's correlation coefficient (r ) = 0.573, p = 0.000; Extract 4: r = 0.675, p = 0.000; Extract 5: r = 0.687, p = 0.000; Extract 6: r = 0.747, p = 0.000; Extract 7: r = 0.733, p = 0.000; Extract 2: r = 0.030, p = 0.809).

The values for the six extracts were merged as shown in Table 11.

Table 11.  All Extracts: Readability and learnability values

There was a significant relationship between readability and learnability for the extracts taken together (r = 0.504, p = 0.000). Passages that are seen as more readable tend to be seen as easy to learn.

Readability and Closeness to French

Extracts 2, 6, and 7 showed a significant relationship between readability and closeness to French (Extract 2: r = 0.274, p = 0.0.27; Extract 6: r = 0.262, p = 0.041; Extract 7: r = 0.561, p = 0.000). When the values for all extracts were merged, we again find a significant relationship between readability and closeness to French (r = 0.173, p = 0.001). This is to be expected as the closer an orthography is to French, the more readable it would be perceived by those who are literate in French.

Learnability and Closeness to French

Individual extracts varied as to whether or not there was a significant relationship between learnability and closeness to French. We find significant results for extracts 4 (r = 0.319, p = 0.013), 5 (r = 0.286, p = 0.031) and 6 (r = 0.352, p = 0.006). Extracts 4 and 5 both use LPT orthography. Extract 6 uses the etymological one. When the values for all extracts were merged, we find a significant relationship between learnability and closeness to French (r = 0.256, p = 0.000). Overall, therefore, in this sample, a spelling system's closeness to French affects perception of its readability.


Respondents were also asked whether the orthography used in the six extracts should be promoted. Sixty-seven percent of informants thought that the etymological spelling should be promoted compared to 41% for grafi legliz and 39% for LPT orthography. In terms of orthography that should not be promoted, we find LPT orthography obtaining the highest percentage (61%) followed by grafi legliz (58%) and the etymological spelling (30%). The breakdown for each extract is given in Table 12 below.

Table 12.  Promotion of the different orthographies
 Should be promoted
 YesNoNot stated
Extract 2 (Etymological)48 (73%)17 (26%)1 (1%)
Extract 3 (Grafi Legliz)29 (44%)36 (55%)1 (1%)
Extract 4 (LPT orthography)28 (42%)38 (58%)0
Extract 5 (LPT orthography)24 (36%)42 (64%)0
Extract 6 (Etymological)41 (62%)23 (35%)2 (2%)
Extract 7 (Grafi Legliz)25 (38%)40 (61%)1 (1%)

Standardisation of Kreol in New Technologies

In the second part of the questionnaire, respondents were asked five open-ended questions that specifically address the issue of language standardisation and use in CMC. We discuss responses in the following sections.

Importance of Same System

Of the 66 informants, 24 (36%) use the same system as their friends in CMC while 40 (61%) do not do so. In this sample, therefore, informants tend to adopt varied spelling conventions when texting and e-mailing. There is no shared standard for most of the informants. However, of the 40 informants who do not conform to a specific orthography, 48% think that it is important to use the same spelling conventions in CMC compared to 40% who do not think that this is the case. More than half of the informants in this sample (61%) believe that it is important that they and their friends use the same orthography for Kreol in CMC. Eleven percent of respondents did not answer this question.

All respondents who think that it is important that they use the same orthography as their friends believe that this will allow for better understanding. Five percent of respondents in this category think that a shared system will foster a sense of belonging and identity among users.

Of the respondents who do not see the importance of having a common orthography for Kreol in CMC, many believe that understanding is more important than a shared system or that Kreol is already easy to read and write. We have such responses as “it is not important as my friends understand my style of writing”; “not important as long as we understand each other”; and “as long as they’ve understood me and I’ve understood them, there is no problem and no need for a shared system.” Kreol is perceived as easy because of its closeness to French. Sixteen percent of respondents claim that it is not important to have a common writing system as there is variation within the different systems proposed –“there are discrepancies between each one style of writing;”“not important as there are at least three types of creole writing”.

Standardising the variety of Kreol used in CMC

Fifty-eight percent of informants believe that the variety Kreol used in CMC should not be standardised while 26% believe that it should. Eleven informants did not answer this question.

Standardisation is not viewed positively in these responses. The differences in the various orthographies proposed and the varieties of Kreol used by separate groups, the fact that a “mixture of languages are used in e-mails and texts” are quoted as reasons against the standardisation of Kreol in CMC. Informants see CMC as a domain where users can adopt their own writing conventions. Having a standard form for Kreol could remove the possibility for the use of different writing styles. One respondent puts it in those terms, it “would be time-consuming at our age to run any creole dictionary and to find the word. It would be a clear disaster!”

Of the 17 informants who supported the standardisation of the variety of Kreol used in CMC, three stated that the standard should be “simple and easy to understand, very easy to learn”/“easy to read and write”, three said that it should look like extract 6 (etymological) or be very close to French, three thought that it should look like extracts 4 and 5 (LPT orthography), one stated that it should look like extract 7 (grafi legliz) and one that it should be “different from French”. There is no consensus on the conventions to be used for Kreol.

Use of Kreol and Ethnicity

Respondents were asked whether some specific ethnic groups were more likely to use Kreol than others. Twenty percent of informants did not answer this question. The majority of respondents (70%) think that Kreol is used by all ethnic groups. Its use was not associated with any particular group: “Creole is our mother-tongue and most of us use it, ethnic group has nothing to do with it as we are all concerned”, “Creole is the most used language amongst Mauritians. It's a national language and is independent of religion or ethnicity”. Kreol is seen as a unifying language that transcends ethnic boundaries. Eleven percent of interviewees thought that some groups are more likely to use Kreol than others: Afro-Mauritians (2 responses), Indo-Mauritians (2 responses). Although two other respondents claimed some groups were more likely to use Kreol than others, they did not state clearly which group they were referring to.

In this sample, therefore, Kreol tends not to be associated with any particular ethnic group. Its widespread use in the community by members of all ethnolinguistic and religious groups makes it an ideal tool for the promotion of a national identity. If Kreol is not associated with any ethnicity, how about the electronic media in which the language is extensively used?

Use of CMC and Ethnicity

The majority of respondents (79%) believed that no ethnic group was more likely to use CMC than others. Twenty percent of informants did not answer this question. Only one respondent thought that some ethnic groups are more likely to use CMC than others. However, he did not state which ethnic group he was referring to.

Just like Kreol, therefore, CMC is used by all ethnic groups. New technologies enable Mauritians of different ethnolinguistic groups to interact with each other through their shared language, Kreol. Since CMC is used by all ethnic groups, it could be used as a tool to promote Kreol among the nation.

Can CMC be Used to Promote the Standardisation of Kreol?

In Part 3, informants were also asked whether the CMC can be used to promote the standardisation of Kreol. Given their extensive use of CMC, respondents were appropriately placed to answer this question. Seventy-nine percent of all respondents believe that CMC can be used to promote a standard spelling system for Kreol while 18% believed that this was not the case. Three percent of informants did not answer Question 4. In this sample, therefore, CMC is seen as a tool in the promotion of Kreol.

Writing Kreol in New Technologies: Some Examples

This section briefly illustrates how Kreol is written in new technologies. Twenty-five postings (totalling 1879 words) from the forum Society on were collected ( Servihoo, a portal especially designed for Mauritius, means “serving you” in Kreol (servi ou) (Ramkhelawon 2005: 46–47). Postings were selected on the basis of the frequency of Kreol words: at least 75% of a selected posting had to be written in Kreol.

The postings focused around Mauritian culture, education, careers, economy, and identity. The aim is not to analyse the contents or structure of the postings but to look briefly at the way in which words are written in Kreol: Are users orienting towards the etymological or the phonemic system? We find both etymological and phonemic conventions used in postings, as shown in the two examples below.

Example 1. Si li vrai qui li meme dimoun li fin change so maniere. Li bon qui li gine changer. Aster la nou capav disquit are li sans aucaine fracas. Mais la maniere qui li introduire ban topic discussion en pay difficile…

(If it is true that she is the same person, then she has changed her ways. It's good that she has changed. Now we can easily discuss with her. But the way she introduces the topics is a bit difficult).

This extract is largely written using etymological conventions. Many of the words are written in exactly the same way as they would in French. For instance, vrai, fracas, mais, and introduire are all written as they would in French. Based on the LPT dictionary, the “Kreol” equivalents for these words would be vre, fraka, me, and introdir. This orthographic system implies two things more or less directly. First, it illustrates how consistency is maintained between the Kreol orthography and the French one. Second, it maintains the traditional linguistic hierarchy and further establishes Kreol as a derivative or an inferior form of French. These etymological conventions can be compared to the phonemic ones, illustrated in Example 2 below.

Example 2. Li enn imiliasion ki nu lim nasional ek nu peyi so langaz ofisiel li langaz ansien kolonizasion ek non pa seki tu morisien koze tulezur. Alor li grantan ki nu kumans balye bann dese kolonial depi nu peyi.

(It is humiliating that our national anthem and official language is the colonial language and not the language that all Mauritians speak everyday. So it's high time that we start getting rid of colonial rubbish from our country.)

The orthographic conventions used here are largely based on phonemic principles and diverge from French. For instance, imiliasion, nasional, ofisiel, langaz would be written as humiliation, national, official, and langage in French. Nasalisation in the words enn and bann is represented by the addition of the final n. Note the use of the letter <k> where we would find <c> in French, e.g., <kolonizasion> for <colonisation> (colonization), <koze> for <cause> (speak). These examples illustrate how <z> replaces <s> in this orthography. In this orthography, the graphemes <k,w,z> are commonly used. This orthography indexes Kreol identity by obscuring the French origins of the words and highlighting their uniqueness. Also, there are no silent letters or other redundant graphemes or symbols.

Table 13 shows the five most commonly used words excluding the personal pronouns mo (1st person singular) and to (2nd person singular). There is only one possible spelling for these two words. We focus on conventions used for the five most commonly used words including words found only in Kreol and words etymologically related to French. It should be noted that generalisations cannot be made from this small sample of words. The aim is to shed some light on how Kreol is written in CMC.

Table 13.  Words most commonly used in postings
WordMeaningPossible spellings (% of times used)Preferred spelling
1. pu/pouModality marker (Mo pu manz–I will eat)pu (83%) pou (17%)pu (phonemic)
2. ki/quiwhich, who, thatki (98%) qui (2%)ki (phonemic)
3. p/peAspect marker (Mo p manz–I am eating)p (92%) pe (8%)p (phonemic)
4. ban/bane/ bann/bannePlurality markerban (17%) bane (38%) bann (29%) Banne (17%)bane (etymological)
5. zot/zottethey/themzot (85%) zotte (15%)zot (phonemic)

Table 13 suggests that at the level of individual words, there is the emergence of spelling conventions. There is a clear preference for one form each of the words listed above (except for the plural marker). Overall, users favour the phonemic spelling over the etymological one. This small selection of postings suggests that there is a discrepancy between what people perceive as being an easy orthography and what they actually use when they write Kreol. Posters orient towards the phonemic rather than the etymological system. It is possible that the choice of orthography is influenced by the identity of writers –e.g., their age, gender, level of education, literacy in French. Further research is therefore needed to expose the relationship between users' identity and their choice of spelling conventions and also the discrepancy between perception and production.


Informants attribute different readability and learnability values to the etymological and phonemic systems used in new technologies. As expected, the etymological system is perceived as more readable and learnable than the phonemic one. This is in line with observations for other Creole-speaking communities where users have acquired literacy in the lexifier (Hinrichs 2004, Mühleisen, 2005). They perceive a Creole orthography that is based on the same conventions as the lexifier easier to read. This is not because the etymological orthography is more systematic –e.g., some of the words in extracts 2 and 6 could have been changed in favour of shorter words –but because the respondents have already learnt to write French and acquired the idiosyncrasies of the French spelling system. Mauritians illiterate in French might find a creole orthography based on phonemic principles and devoid of the inconsistencies of French system easier to read and learn. A further study could assess this claim.

That respondents find the etymological orthography easier to read raises questions regarding the standardisation of Kreol. Should language planners promote an orthography that is based on phonemic principles and that establishes Kreol as a distinct language or one that is based on etymological principles and seems to have support in the community? Or should they cater to those who are illiterate in the lexifier? In the former case, planners would promote an etymological system. In the latter case, a phonemic system would be promoted. This would be systematic and devoid of the inconsistencies of the lexifier and therefore easier to acquire by those illiterate in the lexifier (Siegel, 2005). Do we need to take the number of literates and illiterates into consideration? There are practical as well as identificatory motivations behind the choice of an orthography.

Although users in this study claim that the etymological system is easier to read and learn, Mauritians seem to write Kreol using phonemic conventions on the Net. Indeed, an analysis of some Internet postings shows that users tend to write Kreol using phonemic conventions. It is possible that the phonemic system is preferred only in CMC. That is, CMC is seen as a genre requiring the use of the phonemic orthography. A more quantitative corpus analysis is needed to substantiate this claim and further explain this discrepancy between the orthography that is perceived as easiest to read and learn and the one that is actually used.

This small sample confirms that technology, especially text messaging, has an important role to play in the promotion of Kreol as a written language. Text messages have to be kept short because of the character limit imposed and also, the minute keyboard which makes typing difficult and tedious. It is, therefore, common practice to abbreviate words or use symbols in text messages. Not only does this way of writing save space, but it also saves time and hence, is more convenient. Adhesion to the standard is not the norm in text messages. Since Kreol has no standard orthography (or rather, is thought not to have one), it is up to writers to create their own spelling system. Kreol, it could be claimed, is an ideal language for text messaging: Users can converge on novel codes while texting without ever having to refer to a standard. The very absence of orthographic conventions for creoles appeals to users (Hinrichs, 2004). In fact, standardisation can even be seen “as an obstacle for creative writers who use linguistic variation as a means of expression” (Mühleisen, 2005: 8).

E-mailing differs from texting in that writers do not have to limit their message to a given space and use a small keyboard. In this sense then, e-mails are closer to paper letters than to text messages. However, like text messages, emails tend to be informal (Baron, 1998). Although e-mails are not mentioned as a domain of use of Kreol as often as text messages are, they nevertheless clearly constitute an emerging sphere where the language is used in this sample. Whether in text messages, emails or chats, Kreol offers the writer the convenience of a nonstandard language which can be written in various creative ways. Recent literature has highlighted how written language is creatively adapted to meet the space, time and effort constraints of CMC and also how through technology, new varieties of a language and new identities evolve (Collot & Belmore, 1996; Baron, 2000; Androutsopoulos, 2006b). This means of communication is linked with language creativity, fluidity and “lack of institutional constraints” (Androutsopoulos, 2006a: 429). The use of (non-standard and variable forms of) Kreol, therefore, emphasises the hybrid status of CMC and the creativity and informality associated with this mode of communication.

It should be pointed out that Kreol is not actively promoted in new technologies. In this sense, it differs from minority or lesser-used languages which are consciously promoted by linguists and institutions (Eisenlohr, 2004). In the case of Kreol, we see users themselves rather than those at the policy making end using their language in the new domain of communication. This is more likely to lead to the acceptance and further use of Kreol in other written domains.

In this corpus, the increased usage of Kreol in new technologies is not perceived as leading to the emergence of conventionalised norms. Although users claim that they do not adhere to any norms, discussion of Internet postings and observation of Internet chats confirm that Mauritians have devised their own orthography and interestingly, orient towards some specific phonemic forms. Here again, we note a discrepancy between what users perceive they do and what they actually do.

Moreover, young Mauritians, irrespective of their ethnic identity, use Kreol in the various new forms of communication. None of the respondents claimed using an ethnic language in the new means of communication. When informants use Kreol in electronically mediated communication, they foreground their national identity and background their ethnic one. Through the use of a common language in text messages and e-mails, CMC could lead to the establishment of a shared identity among Mauritian youths. Studies have looked at how diasporic groups use CMC in order to maintain their distinct diasporic identities (see Eisenlohr, 2004). This study suggests that CMC can be used to promote a language and by extension an identity that is common to different ethnolinguistic groups. Further research is needed to assess how different diasporic groups can develop a shared identity through the interaction between language and new technologies.


This study shows that the absence of an officially recognised standard orthography for a language does not deter speakers from using the language in written domains. Although Kreol has not been standardised, young Mauritians of different ethnolinguistic backgrounds write the language in electronically mediated communication.

Previous sociolinguistic research on Mauritius has analysed the way in which ethnic and national identities interact and Mauritians forge their identity based on language ideologies (Eriksen, 1998; Hookoomsing, 2003; Eisenlohr, 2004). Kreol is an important marker of identity among Mauritians in that it unites people of different ethnolinguistic backgrounds. This study shows that people of different origins use Kreol rather than their ancestral language in new media of communication. By using Kreol in text messages and e-mails, informants background their ethnolinguistic identity and foreground their national identity. Kreol, therefore, transcends ethnic and religious barriers not only in spoken but also in written interactions.

This research can be placed within the larger and still embryonic debate on the impact of CMC on language structure and perceptions of identity (Baron, 2000; Crystal, 2001; Georgatopoulou, 2006; Herring & Paolillo, 2006). On the basis of the responses discussed above, we could argue that the new means of communication have the potential to promote the use of Kreol as a written language. A systematic analysis of the variety/varieties of Kreol used in electronically mediated communication will allow us to assess whether increasing usage of the language in such means of communication is leading to the emergence of shared conventions. More quantitative data is therefore needed.

Finally, this study has practical social implications for the standardisation of Kreol, its use in the education system and its promotion at national level –issues that are of immediate interest to Mauritian society generally, and Mauritian language planners specifically. As Kreol has not been officially standardised yet, we have the opportunity to ensure that the standard actually conforms to the varieties used by speakers themselves in new technologies. Involving speakers in language planning decisions is likely to make language policies acceptable to the users.


  • 1

    This study is part of the UK Arts and Humanities Research Council (AHRC) programme Diasporas, Migration and Identities. I am grateful to the AHRC for funding the research for this project (Grant number 119479), conducted while I was working at the University of Edinburgh. I am also grateful to three anonymous reviewers for their comments on an earlier version of this paper. Finally, my thanks go to Jibraan and Saami Carrim for their support and their enthusiasm for this project.

  • 2

    The name of the school has been changed.

  • 3

    The Hindu group also includes students of Tamil and Telugu origins. In Mauritius, this official group is an umbrella category comprising people of Hindu, Tamil, Telugu and Marathi origins.

About the Author

Aaliya Rajah-Carrim is a lecturer in the Department of Linguistics and Phonetics at the University of Leeds. Her research interests include the use of creoles and minority languages in computer-mediated communication, the standardization of creoles, and the interaction between language and identity in multiethnic communities.

Address: Department of Linguistics and Phonetics, The University of Leeds, Leeds LS2 9JT, UK.