This chapter is to outline the relationship between pronunciation and orthography in Mauritian Creole. The orthographic system of Mauritian Creole here is LPT, a system recommended by Ledikasyon pu Travayer which is the spelling system adopted in modern books published in Mauritian Creole. It is also the one employed in Frew��s book. Such relationship discussed here excludes words which are borrowed relatively recently from English or other Indian languages. ��Original�� Mauritian Creole words are the focus, and fortunately foreign words (proper names) are of minority in numbers.

The first part of this chapter will be devoted to the grapheme-phoneme correspondence in Mauritian Creole. Then the second part features syllabication, particularly in connection with the problems associated with the orthographic ��n��.

[Back to top]

Having presented the sounds of Mauritian Creole in the previous chapter, I now move on to discuss how these sounds are represented in writing.

Ideally, for a contact language such as Mauritian Creole, sounds could have simply been represented by their phonetic symbols in order to realize the ��one-sound-one-letter�� perfection, but this is impractical because of problems of and inconvenience in using accents and diacritics in writing, printing, electronic data processing, etc. The LPT orthographic system of Mauritian Creole, instead, opts for ��one-phoneme-one-symbolization��. Digraphs are used, but they are ��no more difficult to learn [�K] than monographs�� (Cassidy, 1993). (cf. Matthews, 2007).

Below are the tables (of oral vowels, nasal vowels and consonants respectively) of the grapheme-phoneme correspondence in Mauritian Creole.

Phoneme

Grapheme

/i/

i

/e/

e

/a/

a

/o/

o

/u/

u

Phoneme

Grapheme

/ɛ̃/

in

/ɑ̃/

an

/ɔ̃/

on

Phoneme

Grapheme

/p/

p

/b/

b

/t/

t

/d/

d

/k/

k

/g/

g

/f/

f

/v/

v

/s/

s

/z/

z

/m/

m

/n/

n (as onset), nn (as coda)

/ɲ/

gn

/ŋ/

ng

/l/

l

/w/

w

/j/

y

/ʁ/

r

At a glance, one might find it unnecessary to have the tables which I have compiled above, since most symbols of phonemes look like how they are represented in orthography. Yes, this is true to some extent that this spelling system is phonetic enough to make it highly learnable. There are, nonetheless, traps to which one has to pay attention.

The biggest problem lies in the fact that /n/ can be either the monograph ��n�� or the digraph ��nn��. If it is ��nn�� in orthography, then undoubtedly it must be /n/. But if it is ��n��, there is confusion of whether it is /n/ as the onset of a syllable or it is to work with the preceding letter to denote a nasal vowel. The solution hinges on syllabification, strategies of which will be discussed later in this chapter.

Other digraphs such as ��ng�� and ��gn�� seem not to pose problems of confusion. A probable reason is that their occurrence in Mauritian Creole is much lower than that of ��n�� and ��nn��. ��ng�� is essentially found in words of English or African origins (Pudaruth, 1972: 46). The use of ��gn�� is not commonplace even in French, the lexifier language of Mauritian Creole. Another proposition may be that ��ng�� and ��gn�� be viewed as the results of assimilation between /n/ and /g/.

[Back to top]

This part attempts to deal with the problem of identifying syllable boundary. Given a phonetic orthography such as that of Mauritian Creole, there would have been little need to outline syllabification if there had not been the confusion of ��n�� discussed above.

First, the syllable structure of Mauritian Creole will briefly be illustrated. Second, how to actually syllabify words will be discussed.

The syllable structure of Mauritian Creole is given by the following formula:

(C)(C)(C)V(C)

1) Nucleus

Each syllable must contain one (and only one) vowel. A Mauritian Creole word can be as simple as just one vowel:

/u/ ��you (formal)��

2) Coda

A syllable may contain no coda or maximally one. All consonants except the bilabial approximant /w/ can be a coda:

/latab/ ��table�� /swaf/ ��thirsty��

/latet/ ��head�� /laplaz/ ��beach��

/disel/ ��salt�� /gaɲ/ ��to earn��

/lataj/ ��waist measurement�� /kon/ ��to know��

/koʁek/ ��well�� /zoʁdi/ ��today��

/feʁ/ ��to do��

It is observed that it is almost always /ʁ/ if a coda occurs word-medially.

3) Onset

The onset of a syllable may contain no consonant (i.e. null onset), one, two or maximally three consonants.

a) With no consonant (the first syllable �V highlighted �V of the following words does not have an onset):

/isi/ ��here�� /ena/ ��have��

/avek/ ��with�� /oʁevwaʁ/ ��goodbye��

/ut/ ��August�� /ɑ̃gle/ ��English��

/ɛ̃vit/ ��to invite�� /ɔ̃zeʁ/ ��eleven o��clock��

b) With one consonant:

/mo/ ��I�� /feʁ/ ��to do��

/latab/ ��table�� /sel/ ��only��

/lanatiʁ/ ��nature�� /jeʁ/ ��yesterday��

/lakaz/ ��house�� /espaɲol/ ��Spanish��

/ʁɑ̃devu/ ��appointment��

c) With a cluster of two consonants:

A cluster of two consonants CC in CCV(C) is possible as an onset. The first C can be one of these consonants: plosives, fricatives, nasals (only /m/ and /n/) and lateral. The second C can be either an approximant or a liquid (/l/ or /ʁ/). The combination is not completely free. Phonotactic rules:

If the first C is a liquid, the second C cannot be a liquid as well.

If the first C is either /s/ or /z/, the second C cannot be /ʁ/.

If the first C is a nasal, the second C cannot be a liquid.

d) With a cluster of three consonants:

An onset cluster of three consonants CCC is also possible. The second and third consonants can only be the sequence /ʁw/. In other words, the cluster is of the form ��Cʁw��, where C can be one of these sounds: plosives or /f/ or /z/.

[Back to top]

In short, the key to syllabification is summarized by the Maximum Onset Principle, which is to ��first make the onset as long as it legitimately can be; then form a legitimate coda�� (Gussenhoven & Jacobs, 1998: 151). The implication is that we prefer onsets than codas, and that if codas ever occur, they should be as short as they can possibly be in a specific language. This principle is not only concerned with the number of onsets and codas - complexity of onsets may also count; see Example 6 below.

Proper syllabification is not only a phonological ideal. It has its practical value since, specifically in this project, it is the key to solving the confusion of ��n�� in orthography as discussed earlier. Here, references to both orthography and pronunciation will be made to clarify how syllables (thus pronunciation) in a Mauritian Creole word should be worked out from its orthography.

Let me illustrate the concept of the Maximum Onset Principle with examples.

Example 1: labutik ��shop��

The way to divide the word labutik into syllables is:

la.bu.tik (A period . denotes syllable boundary. It is also a standard symbol for this purpose in IPA.)

The overall syllable structure is:

CV.CV.CVC [3 onsets, 1 coda]

Referring to the tables of grapheme-phoneme correspondence above, one can easily work out the pronunciation:

la.bu.tik à /la.bu.tik/

Any of the following (and other) ways to divide the word would result in more codas and fewer onsets, which violates the Maximum Onset Principle:

*lab.ut.ik (*CVC.VC.VC) [1 onset, 3 codas]

*lab.u.tik (*CVC.V.CVC) [2 onsets, 2 codas]

*la.but.ik (*CV.CVC.VC) [2 onsets, 2 codas]

*la.bu.ti.k (*CV.CV.CV.C) [3 onsets; but undoubtedly, ��k�� cannot be separated to form a syllable itself, since a syllable must contain one vowel and ��k�� is not a vowel at all.]

Example 2: lanatir ��nature��

Now I start to work with words with ��n��. Here is one: lanatir

The correct way to divide the word lanatir into syllables is:

la.na.tir

The overall syllable structure is:

CV.CV.CVC

Hence the pronunciation should be:

la.na.tir à /la.na.tiʁ/

If someone erroneously divides the word as, say:

*lan.a.tir

there are at least two consequences. First, the Maximum Onset Principle is violated. ��n�� should have been the onset of the second syllable instead of combining with the first ��a�� to form a nasal vowel. Second, due to such violation, now ��n�� makes the nucleus of the first syllable a nasal vowel and the word becomes:

*lan.a.tir à */lɑ̃.a.tiʁ/ (*CV.V.CVC)

The result is two consecutive vowels /ɑ̃a/, an undesirable phonological sequence in many languages, which could have been avoid by proper syllabification.

Example 3: lontan ��long ago��

The word lontan is pronounced with two different nasal vowels but NOT /n/. Syllabification and pronunciation should be as follows:

lon.tan (CV.CV) à /lɔ̃.tɑ̃/

It might be tempting to say ��tan�� as /tan/, but this would imply incorrect syllabification:

*lon.ta.n (*CV.CV.C) à */lɔ̃.ta.n/

��a�� and the final ��n�� would incorrectly be treated as /a/ and /n/ respectively.

At this point, it should start to be clear why knowing what ��n�� is in orthography by proper syllabification is so important.

Example 4: semenn ��week��

In the word semenn, ��nn�� means /n/ and does not combine with the preceding letter ��e�� to form a nasal vowel:

se.menn (CV.CVC) à /se.men/

This is in contrast with ��n�� in lontan of Example 3 above.

Example 5: dimans ��Sunday��

There are two obvious ways to interpret ��n�� in the word dimans. It is either (1) a part of the grapheme ��an�� as a nasal vowel or (2) the phoneme /n/ by itself. Then the two different ways of syllabification are as follows:

(1) di.mans (CV.CVC)

(2) di.mans (*CV.CVCC)

Method (2) is incorrect since it makes the second syllable end with a coda of two consonants, which is forbidden in the syllable structure of Mauritian Creole stipulated above. The pronunciation of dimans, therefore, should be /di.mɑ̃s/.

Example 6: manyer ��way, manners��

Finally I talk about the word manyer in one of the most common greeting phrases Ki manyer?. Again, the problem of syllabification for this word lies in ��n��. Should it belong to the first or second syllable? Let me examine the two possibilities step by step.

First, the two ways to syllabify manyer from orthography are:

(1) man.yer

(2) ma.nyer

which yield these two overall syllable structures:

(1) CV.CVC [2 onsets, 1 coda] (Please note that ��an�� here forms one single grapheme.)

(2) CV.CCVC [2 onsets, 1 coda]

which result in these two pronunciations:

(1) /mɑ̃.jeʁ/

(2) /ma.njeʁ/

On the grounds that (1) and (2) have the same overall syllable structures, the two resultant pronunciations seem to be both valid. No, this is wrong. According to the Maximum Onset Principle, onsets are preferred to codas. Further, as noted above, more complex onsets should be preferred to less complex ones. In this example, (2) gives a more complex onset of the second syllable of manyer than that in (1). Therefore, /ma.njeʁ/ should be the correct pronunciation.

I hope this chapter has adequately explained how to syllabify words in Mauritian Creole and determine how it is pronounced from how it is written, notably in the case with the confusing ��n��.

[Back to top]