Monophthongs

The vowels of New Zealand English are similar to that of other non-rhotic dialects such as Australian English and RP, but with some distinctive variations, which are indicated by the transcriptions for New Zealand vowels in the tables below:^[2]

TRAP–DRESS–KIT chain shift

æ > e̞ > ɪ > ə

• The original short front vowels [æ, e̞, ɪ] (TRAP, DRESS, KIT) have undergone a chain shift to [ɛ, e̝, ə].^[3] Recent acoustic studies featuring both Australian and New Zealand voices show the accents were more similar before World War II and the short front vowels have changed considerably since then as compared to Australian English.^[4] Before the shift, these vowels were pronounced close to the corresponding RP sounds. The stages of the shift are described below.^[5]
  1. TRAP was raised from near-open [æ] to open-mid [ɛ].
  2. DRESS was raised from mid [e̞] to close-mid [e].
  3. KIT [ɪ] was first centralised to [ɪ̈] and then was lowered to [ə], merging with the word-internal allophone of /ə/ as in abbot /ˈɛbət/. This effectively removes the distinction between full and reduced vowels from the dialect as it makes /ə/ a stressable vowel.
  4. The now-close-mid DRESS [e] was further raised to near-close [e̝]. This encroaches on the vowel space of FLEECE.^[6]: 9
  • Realisation of DRESS varies between near-close front [e̝], near-close near-front [ɪ], close-mid front [e], or close-mid near-front [e̠].^[7]
  • Cultivated NZE retains the open pronunciations [æ] for TRAP and [e̞] for DRESS and has a high central KIT ([ɪ̈]).^[1]
  • The difference in frontness and closeness of the KIT vowel ([ɪ̈ ~ ə] in New Zealand, [i] in Australia) has led to a long-running joke between Australians and New Zealanders whereby Australians accuse New Zealanders of saying "fush and chups" for fish and chips^[3] and in turn New Zealanders accuse Australians of saying "feesh and cheeps" in light of Australia's own KIT vowel shift.^[8][9][10]

COMMA

• In the morpheme-final position, the distinction between /ə/ (KIT, COMMA, LETTER) and /a/ (STRUT) is neutralized towards the open /a/ in the word-final position and towards the mid /ə/ elsewhere. For instance, the plural of sofa [ˈsaʉ̯fa] is [ˈsaʉ̯fəz], with the mid /ə/.^[11][12][13] Because of that, the names of the lexical sets COMMA and LETTER are not used in this article.

KIT

• Before the velar nasal, the vowel is much more close and front ([ɪ]) than in other environments. Some speakers also use this variant before /ɡ/ and, less often, before other consonants. It is transcribed with a plain ⟨ə⟩ in this article and so not differentiated from other allophones of /ə/.^[14]
• Initial unstressed KIT is at times as open as STRUT, so that inalterable /ənˈoːltəɹəbəl/ can fall together with unalterable /anˈoːltəɹəbəl/, resulting in a variable phonetic KIT–STRUT merger. This is less common and so it is not transcribed in this article.^[12][13]

FLEECE

• The FLEECE vowel /iː/ may be realised with a slight on-glide when the word is stressed, with /fliːs/ becoming [fləis].^[15] This onglide is increasingly becoming the main way to differentiate FLEECE from DRESS in younger speakers as the latter vowel is a very closed [e̝] and there is a negligible length difference between the FLEECE vowel /iː/ and short vowels.^[6]: 9
• The unstressed close front vowel in happy and video is tense and so it belongs to the /iː/ phoneme: /ˈhɛpiː/, /ˈvədiːaʉ̯/.^[13][16][17]

GOOSE

• The GOOSE vowel /ʉː/ is very central, and may be realised with an on-glide, with /ɡʉːs/ becoming [ɡəʉs].^[15]

NURSE

• The NURSE vowel /øː/ is not only higher and more front than the corresponding RP vowel /ɜː/, but it is also realised with rounded lips, unlike its RP counterpart. John Wells remarks that the surname Turner /ˈtøːnə/ [ˈtøːnɐ] as pronounced by a New Zealander may sound very similar to a German word Töne /ˈtøːnə/ (meaning 'tones').^[18] Possible phonetic realizations include near-close front [ʏː], near-close central [ɵ̝ː], close-mid front [øː], close-mid central [ɵː], mid front [ø̞ː] and open-mid front [œː].^{[19][20][21][7]} It appears that realizations lower than close-mid are more prestigious than those of close-mid height and higher, so that pronunciations of the word nurse such as [nø̞ːs] and [nœːs] are less broad than [nøːs], [nɵːs] etc.^[19][22] Close allophones may overlap with monophthongal realizations of /ʉː/ and there may be a potential or incipient NURSE–GOOSE merger.^[22]

STRUT, START

• STRUT /a/ forms a short-long pair with START /aː/, which means that hut /hat/ contrasts with heart /haːt/ purely by length, like in Australian English. The quality of those vowels is that of retracted cardinal [a]: [a̠, a̠ː], open central [ä, äː], or somewhat higher [æ̠(ː) ~ ɐ(ː)].^[7][23][24]

TRAP–BATH split

• New Zealand English has the TRAP–BATH split: words like dance /daːns/, chance /tʃaːns/, plant /plaːnt/ and grant /ɡɹaːnt/ are pronounced with an /aː/ sound, as in Southern England and South Australia.^[8][25] However, for many decades prior to World War II there existed an almost even split between the pronunciation of dance as /daːns/ or /dɛns/, plant as /plaːnt/ or /plɛnt/, etc.^[26] Can't is also pronounced /kaːnt/ in New Zealand (like Australia but unlike the North American pronunciation /kænt/ with the TRAP vowel). Some older Southland speakers use the TRAP vowel rather than the PALM vowel in dance, chance and castle, so that they are pronounced /dɛns, tʃɛns, ˈkɛsəl/ rather than /daːns, tʃaːns, ˈkaːsəl/.^[27]

THOUGHT

• The THOUGHT vowel may have an off-glide, typically word-finally, turning more /moː/ into [moːə].^[28]

LOT

• The LOT vowel is open-mid, close to [ɞ] and [ɔ].
• The vowel may sometimes be unrounded.^[6]: 10

FOOT

• The FOOT vowel /ʊ/ is close-mid (close to [ɵ]),^[7] and may become centralised, even when stressed, so words like good /ɡʊd/ are pronounced closer to [ɡəd], and could and kid may sound the same, [kʰəd].^[15][29]

Changes before /l/

• Before /l/, /ʉː/ is retracted to [uː], and /e/ is lowered to [ɛ] (see salary–celery merger), yielding a merger with TRAP. These changes make words like too [tʰʉː] sound different from tool [tʰuːl] and leads to Ellen and Alan both being pronounced /ˈɛlən/. Mergers before /l/ may occur between /iː/ and /iə̯/ (as in reel /ɹiːl/ vs real /ɹiə̯l/, the only minimal pair) and /ʊ/ and /ʉː/ (pull /pʊl/ vs pool /pʉːl/).^[25][30]

Māori English

• Māori English has a more fronted and lowered DRESS vowel.^[6]: 16
• The KIT vowel is less central, and is used in unstressed syllables where schwa would be expected^[6]: 16 (due to the merger of KIT and schwa).
• The THOUGHT vowel is lowered compared to General New Zealand English.^[6]: 16
• The GOOSE vowel may be more fronted in Māori English.^[6]: 16
• The NURSE vowel may be more rounded and more fronted.^[6]: 16

Pasifika English

• Features identified as being part of a unique Pasifika English sociolect include a raised KIT vowel, reduced diphthonisation of GOOSE and FLEECE, a lowered DRESS, and for some a retracted and lowered TRAP.^[6]: 17

Diphthongs

Phonetic quality

• On the Cultivated end of the spectrum, the starting points of the fronting-closing diphthongs /æɪ̯/ and /aɪ̯/ are front [æɪ̯] in the first case and central [äɪ̯] or advanced back [ɑ̟ɪ] (both hereafter written with ⟨aɪ̯⟩) in the second case. These are the usual NZE realizations.^[31][32] On the Broad end of the spectrum, they are both retracted, so that /æɪ̯/ acquires a central onset [äɪ̯], whereas the first element of /aɪ̯/ is retracted and rounded to [ɒɪ], sometimes with raising to [ɔɪ] (both hereafter written with ⟨ɒɪ⟩), approaching the CHOICE vowel /oɪ̯/ but without an actual merger. This means that the diphthong [aɪ̯] can stand for either vowel, depending on the variety of NZE. However, unlike the front vowel shift, rounded variants of PRICE are stigmatised, and younger female speakers tend to opt for the conservative variants of those diphthongs even when they exhibit the most advanced variety of the front vowel shift, which leads to the white rabbit [ˌhwaɪ̯t ˈɹɛ̝bət] phenomenon (note the Cultivated [aɪ̯] but Broad [ɛ̝]).^[31]
• The ending points of FACE /æɪ̯/, PRICE /aɪ̯/ and CHOICE /oɪ̯/ vary between close-mid front [e] and close front [i].^[33] In Cultivated NZE, FACE [æe̝] consistently has a higher offset than PRICE [ae̯], much like in General Australian English, but in Broad NZE they normally have the same ending point [e]: [ae̯, ɒe̯].^[1][34] In General NZE, they have been reported to differ as [æe̯] (with a close-mid ending point) vs. [ae̞̯] (with a mid ending point) by one source.^[32] Elsewhere in the article, the offsets of the fronting diphthongs are written with ⟨ɪ⟩ regardless of their precise height, following the way they are usually transcribed in English.
• The onset of /æʊ̯/ is normally raised open front, [æ], whereas its ending point varies between the close back [ʊ] and the close central [ʉ]. Unlike in Australian English, the open-mid back ending point [ɔ] does not occur. In Broad NZE, the starting point is higher, giving [ɛ] or [ɛ̝], whereas the offset is centralized and unrounded to [ə], effectively turning MOUTH into a centring diphthong that encroaches on the Cultivated realization of SQUARE. This [ɛə] realization is gaining ground among younger speakers of the General variety. The Cultivated realization is [äʊ̯] (hereafter written without the diacritic), a glide from the open central position to the close back position, which differs from the General NZE GOAT /aʉ̯/ only by the backness of the second element.^[35][36] According to one source, [aʊ̯] is sometimes also used in General NZE, though more commonly with a somewhat more front onset: [æ̠ʊ̯].^[32]
• The starting point of /aʉ̯/ is [ä], whereas its ending point is close to cardinal [ʉ], making it a glide from STRUT to GOOSE.^[35][37][38] In certain phonetic environments (especially in tonic syllables and in the word no), some speakers unround it to [ɨ], sometimes with additional fronting to [ɪ], making no sound like nigh.^[39] In the Cultivated variety, the onset is mid central and rounded, whereas the ending point is more back: [ɵ̞ʊ̯].^[1]
• The starting points of /iə̯/ and /eə̯/ are identical ([ɪ]) in contemporary NZE. However, conservative speakers distinguish the two diphthongs as [ɪə̯] and [e̞ə̯].^[1]
• Sources do not agree on the exact phonetic realizations of certain NZE diphthongs:
  • The onset of /oɪ̯/ has been variously described as close-mid back [o]^[38] and mid near-back [ö̞],^[32] both overlapping with the allophonic range of THOUGHT /oː/.^[7]
  • The starting point of /ʉə̯/ has been variously described as near-close central [ʉ̞]^[32] and near-close near-back [ʊ].^[38]

CURE

• The CURE diphthong /ʉə̯/ (as in "tour") is becoming rarer,^[40] and tends to be found only following /j/.^[28] Most speakers use either /ʉːə/ or /oː/ instead.^[40]

NEAR–SQUARE merger

• The NEAR–SQUARE merger (of the diphthongs /iə̯/ and /eə̯/) is on the increase, especially since the beginning of the 21st century^[41] so that the phrase that's neither here nor there is pronounced [ˈðɛts niːða ˈhiə̯ noː ˈðiə̯] in General NZE, with here rhyming with there. In Cultivated NZE, the distinction is maintained: [ˈðæts niːða ˈhiə̯ noː ˈðeə̯]. Similarly, beer and bear as well as really and rarely are homophones: [biə̯], [ˈɹiə̯liː].^[3] There is some debate as to the quality of the merged vowel, but the consensus appears to be that it is towards a close variant, [iə̯].^[35][42] The proportion of teenagers showing the merger increased from 16% in 1983 to 80% in 1999.^[43] The merger is nearly complete, with most younger speakers being unable to tell the two diphthongs apart.^[6]: 13 As the merger is not yet fully complete, it is transcribed only in phonetic transcription, whereas in phonemic transcription the distinction is maintained: /ˈðɛts niːða ˈhiə̯ noː ˈðeə̯/, etc.

Changes before /l/

• Before /l/, /aʉ̯/ becomes [ɒʊ̯],^[32] making go [ɡaʉ̯] sound different to goal [ɡɒʊ̯ɫ]. This vowel change may lead to a merger with LOT (/ɒ/) (doll [dɒɫ] vs dole [dɒʊ̯ɫ]),^[44] especially when the /l/ is vocalised. This has been labelled the GOLD lexical set by Bauer et al. (2007:98).

Transcriptions

Sources differ in the way they transcribe New Zealand English. The differences are listed below. The traditional phonemic orthography for the Received Pronunciation as well as the reformed phonemic orthographies for Australian and General South African English have been added for the sake of comparison.