Variation in Fijian: Boumaa and Standard Fijian compared

This essay was originally written as ‘A comparative analysis of Boumaa and Standard Fijian‘ in Semester 1 of 2019 for LING2001, Languages of Australia and the Pacific, at the University of Newcastle. The body of the essay is unrevised.


Fijian is a set of two languages in the Central Pacific sub-group of the Austronesian language family. It’s something of an outlier amongst other Central Pacific languages in that the dialects on the extremities of the island-group aren’t mutually intelligible, whereas most of the sub-group consist of a language for each island or island-group (Lynch, Ross & Crowley, 2002, p.6) (John, Malcolm, & Terry Crowley, 2002, p. 6).  

Boumaa is a variety of Fijian spoken primarily on the island of Taveuni off the coast of Vanua Levu, in immediate proximity with the more prominent varieties Ca’audrove (Cakaudrove in Standard Fijian) and Waini’eli (Wainikeli in Standard Fijian). Taveuni is quite geographically removed from the island of Bau from where Standard Fijian originally emerged, but the Eastern Fijian dialect chain, of which Boumaa and Standard Fijian are both part, distinctly crosses the archipelago. 

Phonetics & Phonology 


There’s not much to be said about the difference in vowels between Boumaa and Standard Fijian. Both use the same set of cardinal vowels, [i, e, a, o, u], and both have phonemic length as well as diphthongs beginning with a non-high vowel and finishing with a high vowel, with the stress on the beginning vowel (Dixon, 1988, p.14-5)(Schütz, 2014, p.4-5). 

Apart from some other outlying examples including loanwords, Schütz (2014) lists iu as part of the Standard Fijian vowel inventory alongside the rest, without giving examples in his chapter on vowels, whereas Dixon (1988) gives it as only identifiable in four utterances in Boumaa including a place name, where the syllabic peak is on the second vowel rather than the first. This appears to be somewhere between a vowel and a palatal approximant, but I shan’t go into the phonetic nitty gritty. There could be some difference in this, but by Schütz’s lack of example, I wouldn’t conclude that. 

Schütz determines the /o/ phoneme as closer in Standard Fijian to [o] than [ɔ], but Dixon describes the phoneme just “covering both” in Boumaa. Similarly Schütz refers to the /e/ phoneme as being commonly heard in Standard Fijian as [ɛ] rather than [e], but Dixon again describes the phoneme as covering both these sounds. 


Many of the same consonantal phonemes are shared between the two varieties. It’s likely that some of the variations that can be seen between Shültz and Dixon are only difference in transcription, but there is at least one noteworthy consonantal difference. 

Whilst voiced plosives are expressed by Schütz as a nasal-initial consonant cluster, and by Dixon as a prenasalised consonant, these are basically the same thing. The two put /w/ in starkly different places of articulation, but they’re both correct because /w/ is the doubly articulated voiced labio-velar approximant. To my knowledge, the inverted circumflex or “caron” that Schütz uses above the alveolar trill is representative of a rising tone, but he’s also got nasalisation there in a sort of range so I’m not really sure what’s going on there. Close enough. 

Schütz uses some kinda outdated symbols for affricates and also denotes an initial nasal in the voiced affricate, which Dixon doesn’t note in the same way, but later refers to a prenasalised voiced affricate as an allophone of the /d/ phoneme before /i/. Though in both cases, the use appears fairly marginal. 

In both varieties, the /f/ phoneme is primarily if not only used in loanwords, though understandably Standard Fijian has a loanword inventory from a lot more languages (Schütz, 2014). 

The difference in the /y/ phoneme looks a little weird in Schütz’s transcription of Standard Fijian. The diacritic is an inverted breve below, which is used in the IPA to denote a sound as non-syllabic. As such, Schütz is transcribing an approximant or semi-vowel in the mid-low front region of the vowel space, whereas Dixon transcribes it as a simple palatal approximant [j], which is in the high-front region of the vowel space. For the few words where this phoneme does appear, from Schütz’s transcription words like <toyovu> would be pronounced [toɛ◌̯ovu] in Standard Fijian whereas from to Dixon’s transcription would be pronounced [tojovu] in Boumaa. This is a fairly minute difference and as such it’s possible that the approximant in Boumaa was just a case of the Fijian phoneme and Dixon’s English phoneme being considered allophones of each other, or the difference simply not being noticed, but this can’t really be known without further research into Boumaa. 

The most distinct difference is the use of glottal stops in Boumaa, which only appear in Standard Fijian in loanwords from other Eastern Fijian varieties such as Boumaa. For all terms in Standard Fijian that contain instances of a single /k/, the phoneme and even grapheme is changed to that of a glottal stop in Boumaa such as the verb for “steal” which in Standard Fijian is butako, and in Boumaa is buta’o. There are however exceptions to this. In Boumaa most English loanwords retain a velar stop such as motokaa, but older loans such as tapa’o (“tobacco”) have undergone that shift. Similarly, more recent loans from Standard Fijian are pronounced with either sound. The Standard Fijian mataka for “morning” is sometimes pronounced mata’a, but has pretty well replace the original Boumaa word for morning, saubogi (Dixon, 1988, p.13). 

Whilst Standard Fijian only has occurrences of the /p/ phoneme in loanwords, and many loanwords have at least partially been reanalysed to the /v/ phoneme (Schütz, 2014, p.353-4), Boumaa has at least a dozen indigenous words that contain the /p/ phoneme both in initial position and otherwise (Dixon, 1988, p.4, 364-6). The Boumaa cognates of numerous Standard Fijian words with initial /n/ have that /n/ dropped (Dixon, 1988, p.4). 

According to Schütz and Dixon’s grammars, Boumaa and Standard Fijian have an identical basic syllable structure, though Dixon describes more detail regarding vowel use, particularly across word boundaries and reduplication boundaries (Dixon, 1988, p.15-16). Whether the occurrence of words such as <iloilo> (“glass”) being pronounced [ilo.ilo] with separate monophthongs across the syllable boundary is a specific feature of Boumaa Fijian can’t be reasonably determined with the limited data provided by Schütz (2014). 

Prosody & Phonotactics 

A prosodic comparison of Boumaa and Standard Fijian proves even more difficult, again using Dixon and Schültz as primary sources, as whilst Dixon covers the matter of phonetic stress (1988, p.16-18), Schültz distinctly and deliberately approaches the matter of Standard Fijian prosody phonologically, referring instead to “accent” and “measures” (2014, p. 5-6, 370-390). Regardless of this, some useful comparisons can be drawn. Both varieties are more or less in line with the simplified general stress of Oceanic languages, that primary stress falls on the penultimate syllable (Lynch et al., 2002, p.35). Though, with more nuance, Boumaa Fijian’s stress is described in terms of moraic value rather than purely syllabic value, primary stress being on the syllable with the penultimate mora, and secondary stress being on the “syllables containing the fourth and sixth moras from the end” (Dixon, 1988, p.17). This is reminiscent of what Schütz calls the “Polynesian accent myth”, but as mentioned, is more specifically regarding moraic value than the syllabic value of that accent myth and the 1995 example that Schütz criticises (2014, p.382). The system of stress of Standard Fijian appears at first glance to fit the definition of Boumaa Fijian stress, but is somewhat convoluted by Schütz’s more phonological approach. 

Standard Fijian features shortening of long vowels, both monophthongs and diphthongs, before an unaccented syllable, leading to terms like <beca> [ˈbeða] from <be> [ˈbeː] (Schütz, 2014, p.374). Such shortening is reflected in Boumaa Fijian (Dixon, 1988, p. 26-7), but not diphthongs. It is apparent by acoustic measurements that diphthongs in the same context in Standard Fijian are also shortened (Condax as cited in Schütz, 2014, p.374), suggesting that the diphthongs that carry stress are more distinctly utilised as a single phonemic unit, in contrast with its following unstressed syllable, than their isolated counterparts (Schütz, 2014, p.373-4). This seems potentially incongruent with Boumaa Fijian’s stress system, but it could be reasonably assumed that the lack of evidence of this phenomenon in Dixon’s grammar (1988) could be from a lack of acoustic approach and data set, though Schütz draws on an auditory impression from a text published one year after Dixon’s grammar, so whilst this might be a distinct difference between the two varieties, a conclusion cannot be drawn without analysis of a depth beyond the scope of this report. 

In phonotactic summary, Boumaa Fijian has a distinct monophthong differentiation between syllable boundaries and shortens long monophthongs that occur before unaccented syllables (Dixon, 1988, p.15-6, 26-7), whereas Standard Fijian shortens both long monophthongs and diphthongs that occur before unaccented syllables (Schütz, 2014, p.374). In prosodic summary, Boumaa Fijian has a pattern of primary stress on the syllable containing the penultimate mora with secondary stress on the syllables containing the fourth and sixth from last morae, whereas Standard Fijian has a pattern of primary stress on the final syllable if it contains a long vowel, or on the penultimate syllable if the final syllable contains a short vowel. This, in combination with the aforementioned Standard Fijian phonotactic feature, appears to imply that stressed vowels in Standard Fijian are always either long and word-final, or short and not penultimate. It’s important to note that these compared features are not necessarily mutually exclusive, lest one draw conclusions about each variety that is not necessarily true. As such with the limited information available, these can only be considered as possible differences. 


Perhaps not a particularly significant difference, but it is worth noting, particularly with respect to the difference in phonology and the intrusiveness of Standard Fijian in the social context of Boumaa Fijian, that the orthography that was developed for what has now become Standard Fijian, is used in a vast majority for Standard Fijian rather than Boumaa Fijian or the other surrounding dialects (Dixon, 1988). This orthography also does not accurately reflect the phonetics or phonology of the Fijian varieties on and around Taveuni that use glottal stops (Dixon, 1988, p.4).  


A particularly distinct and noteworthy difference between Standard Fijian and Boumaa Fijian is their inventory and use of deictics. These included different demonstratives and deictic verbs (Dixon, 1988, p.58). 


Fijian has three kinds of demonstratives: 

  • Proximal (near first person), for that which is near the speaker. Roughly equivalent to English this/here. 
  • Near-mid distal (near second person), for that which is away from the speaker, and often near the addressee. Partially equivalent to English that/there. 
  • Far distal (near third person). It could be reasonably inferred that it is distinctly away from both the speaker and addressee. Partially equivalent to English that/there. 

Standard Fijian uses two series of these kinds for different contexts, whereas Boumaa Fijian only uses one series. The Standard Fijian term for a far distal demonstrative that is used in contexts apart from after a preposition, (o)yaa is shared with the Boumaa Fijian term for a near-mid distal demonstrative, yaa, and the Boumaa Fijian far distal demonstrative, mayaa appears to be a combination of the Ca’audrove and Standard Fijian far distal demonstratives, maa and (o)yaa respectively. Ca’audrove Fijian has a split system similar to Standard Fijian for proximal and near-mid distal demonstratives, but just the one far distal demonstrative. This kind of dialect-mixing is reflective of the “bewildering pace” of code-switching between various demonstratives of these three and other varieties of Fijian that undergoes amongst speakers of Boumaa Fijian (Dixon, 1988, p.58-9). These demonstratives are used for reference to spatial proximity/distance as well as spatial proximity/distance and anaphoric use, or reference to the subject being spoken/written about (Dixon, 1988, p.58-61). Standard Fijian uses its second set of demonstratives for anaphoric use (Schütz, 2014, p.242-4). 

Deictic Verbs 

Boumaa and Ca’audrove Fijian have a distinct set of deictic verbs, ena and ene(ii), that can be used with or without an appropriate demonstrative to express a “do like this/that” or “look like this/that” verb function. Standard Fijian does not appear to have an equivalent term so is most closely substituted by vaa-qoo and vaka-yaa/vaa-yaa (Dixon, 1988, p.61). These terms will likely have developed separately from the varieties found on and around Viti Levu, such as Standard or Bauan Fijian, and with the equivalence of the Standard Fijian /k/ and the Boumaa/Ca’audrove /’/ phonemes, it would be understandable for ena not to enter the lexicon with the noun and Nroot kena that exists in Standard Fijian (Gatty, 2009) and possibly already did exist in Bauan Fijian for some time. 

Boumaa Fijian also has the deictic transitive verbs va’aa and va’aa-ta’ina meaning “relating to” or “like” (Dixon, 1988, p.61). Standard Fijian of course has the phonemically corresponding vakaa which can, amongst many other uses (Schütz, 2014, p.56-66), be used to mean the same (Gatty, 2009), but does not have an equivalent *vakaa-takina. Standard Fijian does have the word takina, phonemically equivalent to ta’ina in va’aa-ta’ina, the meanings of which are that of a part of the body or that of a social or familial relationship (Gatty, 2009), and is thus a similarly categorical relating term. This suggest that the two terms could be cognate in some fashion, but to conclude as such on this limited basis would be presumptuous. 

Deictic Nouns 

Boumaa Fijian also has a deictic noun amu, meaning “that sort of” or “that one”. It can be used anaphorically or a restriction of reference (Dixon, 1988, p.62). Standard Fijian does not have a phonetically equivalent *kamu (Gatty, 2009), nor seemingly any other semantically equivalent term (Schütz, 2014). 

Grammatical Voice 

Passive Voice 

Dixon describes the passive voice as “[p]erhaps the most significant grammatical difference [between Boumaa and Standard Fijian” (Dixon, 1988, p.4), in that Boumaa Fijian verbs with polysyllabic suffixes are trisyllabic in active voice, as in the –ta’ina in au liu-ta’ina (“I commanded him/her/it”) and the –ta’ini in au liu-ta’ini Jone (“I commanded John”), but disyllabic in passive voice, as in the ta’i in liu-ta’i (“commanded”), whereas Standard Fijian verbs with polysyllabic suffixes are disyllabic both in active voice, as in caka and caki, and in passive voice, as in caki (Dixon, 1988, p.47). 

Stative Voice 

In contrast with Dixon, Schütz asserts that the grammatical voice of Fijian often identified as passive is more accurately described as stative (2014, p.95). However, the examples and translations given by Schütz as being stative could reasonably also be interpreted as passive. bulu (“She is buried”) appears that it could mean “She is undergoing a change from not being under a mound of dirt to under a mound of dirt”, such as if a landslide were to fall on “her”, which would be stative; or it could mean “She is being affected by people who are burying her”, such as if she were having sand shovelled onto her at the beach, which would be passive. Without much first-hand experience of the language it is exceedingly difficult to infer whether implications of an agent are present in such examples, but it is worth noting that the two examples given by Dixon regarding Passive voice each overtly do or don’t imply an agent. 


The only significant phonemic difference is a pulmonic egressive stop moving between the velar and glottal places of articulation in identical distribution. With more data some minor prosodic and phonotactic differences might be determined. Boumaa Fijian has some more deictic terminology in its lexicon, but nothing that a Standard Fijian speaker couldn’t reasonably work around; and the passive/stative voice has a small but distinct morphological difference. It’s vaguely possible, but very unlikely that one variety has a distinctly passive voice and the other has a distinctly stative voice, but that ought not to be wholly ruled out without further exploration. Codeswitching even occurs between Boumaa and Standard Fijian (amongst other varieties) in the use of demonstratives. 

There are more factors to consider, but from the above it is quite apparent that Boumaa and Standard Fijian are two varieties of the same one language.


Condax, I. D. (1979). Syllable durations and phonological groups. Journal of the Acoustical Society of America, 66(S1) 

Dixon, R. M. (1988). A Grammar of Boumaa Fijian. Chicago: University of Chicago Press. 

Gatty, R. (2009). Fijian-English Dictionary with notes on Fijian culture and natural history. Suva, Fiji: Ronald Gatty. Accessed from: 

Lynch, J., Ross, M., and Crowley, T. (2002). The Oceanic Languages. Richmond, Surrey: Curzon Press. 

Schütz, A. J. (2014). Fijian Reference Grammar. Honolulu: PacificVoices Press. 

Note: Gatty (2009) appears to be self-published, but is recommended by the Cornell University Press as “the most up-to-date lexicographic source for the language, a reliable, practical guide that includes helpful notes on word usage and Fijian culture.” (

Feature Image by Nicolas Weldingh on Unsplash

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: