Absence of yod in non-primary stressed /Cju/ syllables

The deletion of the palatal glide or approximant [j] (commonly referred to as “yod”) in primary stressed /Cju/ syllables is found in varying degrees in the “inner circle” Englishes (Kachru 1985: 12) and is generally the result of various historical processes. Most of these varieties (e.g. Cockney, Estuary, General American, Australian, New Zealand, South African English etc.) occupy positions between that of conservative Received Pronunciation, which has the least amount of yod-deletion, and East Anglian English, which deletes yod in all phonological environments (Wells 1982). The most common type of yod-deletion is found after alveolars and dentals, e.g. in General American English (Wells 1982) and in Cockney (Wells 1982). In these two varieties, the deletion only occurs after alveolars and dentals – not after labials or velars. Therefore, items such as music and cute are never realized as *[muzik] or *[kut]. With the exception of East Anglian English, yod-deletion in non-primary stressed syllables (no matter what the preceding consonant) is not usually found in any “inner circle” variety.

With the exception of those lexical items in which yod has been historically deleted in most varieties of English (e.g. rude, blue), another type of yod-deletion occurs in Fiji English and is probably the most prominent phonological feature across the whole spectrum of its speakers. It involves the absence of yod in non-primary stressed syllables, not only after alveolars and dentals (with no evidence of [tj] > [ʧ] or [dj] > [ʤ] coalescence), but also after labials and velars, e.g. regular > ['rεgula] , stimulate > ['stimule:t] , annual > ['εnuɐl], situation > [situ'e:ʃɐn] , popular > ['pɔpula] , educate > ['εduke:t] , fabulous > ['fεbuləs] , occupation > [ɔku'pe:ʃɐn].

Tent (2001) conducted a detailed quantitative analysis of this phenomenon and found that although it is dynamic and complex, the absence of yod in non-primary stressed /Cju/ syllables was primarily a characteristic of L2 Fiji English. However, it had gradually evolved into a phonological shibboleth of many, if not most, L1 Fiji English speakers. The distinct clines in yodless pronunciation in terms of age, gender and level of education indicate that a change is in progress. The younger the speaker is, the more yodless is the pronunciation, while the more educated the speaker is, the more yod is used (or retained) in this particular phonological environment.

There is also a tendency for females to favor a yod pronunciation, which concurs with the findings of most other social dialect studies which report that women tend to use a more standard or prestige pronunciation. The tendency for males to favor yodless pronunciations, especially the younger ones, suggests that the phenomenon is a marker of covert prestige, maleness, and group identity (i.e. being speakers of Fiji English). The desire to identify with the local community is strong and is manifested linguistically. The reasons for the retention of yodless /Cju/ syllables, and indeed its increase among young Fiji English speakers, may well be because its speakers do not wish to alienate themselves from those within their own speech community.

Educated speakers in Tent’s study also have yodless /Cju/ syllables, but to a lesser extent. This suggests that these speakers may be more linguistically sensitive and aspire to speak Modified Fiji English. More empirically based sociolinguistic research in this area is required to determine how strongly pressure to conform to the local norm is felt by the various sub-groups of Fiji English speakers, and to what degree this depends on how closely enmeshed the speaker is in the community.

On the other hand, some speakers realize blew as [bliu] and flew as [fliu]. For those speakers, blue and blew, and flu and flew are homophonous. This yod insertion may be more common among Indo-Fijians, but it is also present in the pronunciation of some Fijians. It may be an over-generalization of the pronunciation of orthographic <-ew> in general (e.g. few [fju], new [nju]), or more specifically in the past tense of strong verbs (e.g. knew [nju]). Some informants seem to think that this is a teacher-induced error. These observations are, however, based on data from only about thirty informants, and confirmation would require more systematic empirical study.