← All chapters

§3 Phonology

Pine has a rich phonology, with a large vowel and consonant inventory and complex phonological processes, namely diaphony (changes in the quality of the stressed coda) and apophony (changes in the quality of the stressed nucleus), which both serve morphosyntactic purposes and are central to Pine grammar. Pine has both back and front-rounded vowels, which can be either lax or tense and distinguishes two lengths: short and long. One distinctive feature of Pine phonology is a type of semivowel harmony of /w/ which appears both as [w] and [ɥ] depending on its environment, which often proves difficult for non-native speakers. Pine has fixed stress, which always occurs on the first syllable of a root. However, as Pine makes use of a number of prefixes and stress is not indicated through orthography, one must have some level of familiarity with the various affixes to correctly assign stress in many cases.

Below is a list of phonological terms and their definition that are specific to Pine phonology:

§3.1 Vowels

Pine has eight vowel phonemes: /a ɑ o u ø y e i/ that yield 26 different vowels: eight short lax vowels [a~ɑ ɔ ʊ œ ʏ ɛ ɪ] <a o u ů y e i>, seven long lax vowels [ɑː ɔː ʊː œː ʏː ɛː ɪː] <a o u ů y e i>, eight geminate vowels [æː ɑː oː uː øː yː eː iː] <ả oa ỏ ủ eů ỷ ẻ ỉ> and three diphthongised vowels <uo ůa ůe>. Long tense vowels, called geminate vowels, are written with a ˀ (called nįottahįỏ) <ả ỏ ủ ẻ ỉ ỷ>, except for /ɑː/ [ɑː] <oa> and /øː/ [øː] <eů>, which are written as digraphs. Geminate vowels are always long and tense, regardless of their environment. The three high rounded vowels /y ø u/ have a geminated form in <ỷ eů ủ> [yː øː uː] and diphthongised forms in <ůe ůa uo>.

The vowels /a ɑ/ have converged when lax. They are only distinguished when geminated, where /a/ is geminated to [æː] <ả> and /ɑ/ is geminated to [ɑː] <oa>.

Lax vowels can either be short or long. Short lax vowels occur in closed and word-final syllables, except for <a>, which is always [ɑ] word finally and in open syllables, and [a] <a> only occurs in closed syllables. A closed syllable has a complex coda – any non-simple consonant will close a syllable, including geminated consonants, preaspirated consonants such as [ʰp ʰt ʰk] <pp tt kk> and labialised consonants, or a word-final syllable that ends in a consonant, for example -at [-at]. Long vowels occur in open stressed syllables and in open penultimate syllables, and in most dialects a geminate consonant cause the syllable to be close, for example in ƛukƛuggut ‘perfect, fitting’, which is more commonly pronounced as [ˈt͡ɬʊkːt͡ɬʊkːʊt], not as [ˈt͡ɬʊkːt͡ɬʊːkːʊt], although the latter is sometimes heard in western dialects, where penultimate lax vowels can become long if in a lytic syllable. In simieṡkinuƛami ‘(the fire) starts to burn well, make large flames’, the prefix si- is unstressed and short, -mieṡk- has a short diphthong, -inuƛ- are both open syllables but both vowels are short as they are not in penultimate position, and -ami has a long -a- as it is the penultimate syllable, is lax and in an open syllable, giving [sɪˈmɪɛ̯ʃːkɪnʊt͡ɬɑːmɪ]. Stressed monosyllabic words with a short lax vowel become long, especially forms of the copula such as įus [ˈjʊːs] ‘should be’ or nin [ˈnɪːn] ‘(which) is’.

Diphthongised vowels differ from true diphthongs in that their pronunciation depends on syllable structure and labialisation. They go from a lax-glide combination in closed syllables ([ʊu̯ œy̯ ʏy̯]) to a tense-glide combination in open or lytic syllables or when labialised ([uʊ̯~uɔ̯ øœ̯ yœ̯]). Additionally, the diphthongised vowels <vuo vůe> can be found as either [wuʊ̯ ɥyœ̯] or [woʊ̯ ɥyʏ̯].

In eastern dialects, labialised diphthongised vowels are often lowered to [ᶣɛ ᶣæ ʷɔ] <vůe vůa vuo>.

The vowels of Pine are therefore contrasted by short and long lax or tense, front or back, rounded or unrounded and close or open articulation, in addition to the three rounded diphthongised vowels <ůe ůa uo>.

Pine vowels have great symmetry in all cases except for /a/, which has two geminates, the front [æː] <ả> and the back [ɑː] or alternatively [ɒː] <oa>. In certain eastern dialects, long <a> and <oa> are separate phonemes which are distinguished by lip rounding, while in most other dialects they are no longer distinguished, and both have converged to [ɑː], which is the transcription found in this book. However, they behave differently with regard to apophony, which is why they are distinguished in orthography. The non-geminate /æ/ is only found in the sequence <ai, ůai~vai> [æɪ̯ ɥæɪ̯] in modern Pine, or in the case of diphthong reduction due to dentolabialization or sibilant fricativization with preaspirated clusters, as in gvaikkevyt [ˈgᶣæɪ̯ʰkɛːɥːʏt] ⟫ [gᶣæ(ɪ̯)ɕːkɛːɥʏt] ‘universe’.

§3.1.1 Secondary Vowel Lengthening & Secondary Stress

Lax vowels found in non-stressed syllables become lengthened in certain environments. Namely, lax vowels found in penultimate syllables become long if they are not directly found after a long lax stressed vowel. This phenomenon is called secondary vowel lengthening. This means that any word with three syllables or more will have a long penultimate vowel if it is lax and in an open syllable, unless the preceding stressed vowel is also lax and long. Syllables containing a long lax vowel are called heavy syllables (H), and syllables lacking a long lax vowel are called light syllables (L) and the general pattern in Pine is either H–L (trochee) or L–L (dibrach) if the stressed syllable is light for bisyllabic nouns and H–L–L (dactyl) or L–H–L (amphibrach) if the word contains three syllables, or in the case of a stressed syllable containing a light syllable, L–L–L (tribrach).

Lengthening a lax vowel immediately following the stressed syllable (H–H–L, antibacchius) is nearly exclusively heard in western dialects and standardised language nowadays, whereas eastern dialect speakers and standard language registers strongly favour H–L–L. It remains one of the most instantaneously recognizable traits of western dialects. For example, tṇegema ‘it sees her/him’ would be pronounced [ˈtn̥ɛːgɛːmɑ] in some western dialects.

Secondary stress occurs on the first syllable of the last foot (either a bisyllabic or trisyllabic unit) of a word. Secondary stress and secondary lengthening occur in the same syllable. For that reason, syllabification of a word tends to favour a trisyllabic initial foot and subsequent bisyllabic feet so that the penultimate vowel can have secondary stress, and lengthening if it applies.

In the word 6 syllable word keutseminttilitṡin 'we_incl would come have along', stress falls in -mintt- (from mindat 'to come'), but the first syllable of the first person inclusive plural -itṡin bears secondary stress and length as it is an open and penultimate syllable not directly preceded by a stressed syllable.

The word is analyzed as L–L–H–L–H–L or keu|tse|mint|til|itṡ|in, breaking down the word as a dibrach (L–L) and two trochees (H–L), which results in secondary stress and lengthening on -iːtṡin.

There is a corresponding phenomenon to secondary lengthening in penultimate syllables found in syllables that precede the stressed syllable called secondary peninitial lengthening . Whereas penultimate indicates the second to last syllable, penitial means the second to first syllable, and secondary lax vowel lengthening also occurs there. Some speakers have the same heavy-light pattern in prefixes. Penitial lengthening does not occur if the peninitial syllable is a non-initial (virtually always the second syllable) or a prefix. Additionally, the kinetic initials si(s) and geg-/gi- are commonly found without secondary lengthening, especially in eastern dialects.

§3.1.2 Diphthongs

Pine distinguishes 16 different true diphthongs, excluding the three diphthongised monophthongs <ůe ůa uo>. The diphthongs <ey, ay, iů, eu, au, iu> can be thought of as being post-rounded, or consisting of a lax vowel followed by either [u̯] or [y̯], such that they are found as [ɛy̯ æy̯ ɪy̯ ɛu̯ ɑu̯ ɪu̯]. Front post-rounded diphthongs are typically pronounced with lip protrusion, such that <iů> can be narrowly described as [ɪy̯͗], while back post-rounded diphthongs usually have lip compression, such that <iu> can be narrowly described as [ɪu̯ᵝ].

The other non-rounded diphthongs are <ei, ůi, ui, ua, oi> [ɛɪ̯ œɪ̯ ʊɪ̯ ʊɑ̯ ɔɪ̯~ɑɪ̯], of which <oi> can be heard as [ɔɪ̯] or [ɑɪ̯], although the latter is often considered more standard or correct, and it contrasts with <ai> [æɪ̯]. The diphthong <ua> [ʊɑ̯] is not commonly found in stressed syllables. In eastern dialects, the diphthong <ui> is pronounced [ʏɪ̯] in open syllables before a front vowel, such that kvuiniḳ ‘oak lichen’ can be pronounced [ˈkʷʊɪ̯nɪkʷ] or [ˈkᶣʏɪ̯nɪkʷ] in certain eastern dialects.

Word-final <ia> may be heard as either [ɪɑ̯] or [ɪ̯ɑ], or even [jɑ], or [ɪa̯] before a word-final consonant.

§3.1.3 Semivowels

The semivowel /w/ <v> has four possible realizations in Pine: [f~ʋ w ɥ]. Directly following -ḷ- or -ṛ-, it is pronounced as [f] or [ʋ], such that the sequences <ḷv ṛv> are pronounced [ɬːf, r̥ːf] or [ɬːʋ, r̥ːʋ] or [lːʋ, rːʋ]. Additionally, the word initialy sequence <h+v> is realised as [p] <p>, such that hulġġu 'hut' ⟫ půelgimyt ‘in the hut’ (where the nucleus -u- undergoes apophony to -vůe-, which causes h-vůe- to become půe-) Tuġvut distinguishes these two types of <p>, whereas Teulgi does not. When geminated, <vv> is pronounced [ʋː] in most dialects, but some western dialects have conserved lip rounding distinction as [wː] or [ɥː].

In all other cases, there are two contrasting pronunciations of /w/ <v>: [w] or [ɥ]. Intervocalically, front vowels and diphthongs beginning with a front vowel (ả e i y ů ai) cause /w/ to be front-rounded to [ɥ] , while /w/ appears as [w] before back vowels (a o u). In the case of the diphthongs <eu iu>, the semivowel /w/ is most commonly realised as [w], as in seuvittit [ˈsɛu̯wɪʰtɪt] ‘into rivers’. A distinguishing feature of eastern dialects is that word initially or in stressed onset position, /w/ <v> is pronounced [w] before <e i> unless it occurs directly before <ů>. This is also true when /w/ is found after non-plosives, in which case [ʋ] is common.

One exception to the allophony of /w/ is that the sequence [Vu̯ɥV] is realised as [Vu̯wV]. That is, when /w/ is expected to be pronounced as [ɥ] but is directly preceded by a diphthong ending in [u̯] (<au uo iu eu>), it is instead realised as [w], for example in seuvittit [ˈsɛu̯wɪʰtɪt] ‘river (illat.pl)’. This is called progressive glide metaphony. Some speakers, especially in northeastern dialects, go the other way and instead change the nature of the preceding diphthong such that <au uo iu eu> become <ay ůe iů ey> to maintain [ɥ]. This phenomenon is termed regressive glide metaphony.

Directly following a consonant, the semivowel /w/ is usually transcribed as [ʷ] or [ᶣ], because labialised consonants that are geminated as a whole, such that both the length of the consonant and the glide is lengthed as a single unit. Nonetheless, labialised consonants are counted as a cluster when it comes to their effects on syllable structure because they are either preaspirated or geminated – they are never short intervocalically.

before fv and after <au uo iu eu>

The letter <v> cannot appear after <n> or <m> in Pine orthography, or after <l> word initially. A handful of words have a glide in that position, but only <ůei, ůey, ůai, ůay, uoa> occur, corresponds to <vei, vey, vai, vay, vau>. Tuġvut uses <v> in such cases, although alternative forms following Teulgi orthography are also accepted

§3.1.3.1 Diphthongs & Consonant Labialization

Rhizal apophony can change the environment in which a word-initial onset consonant finds itself. Special rules apply in cases where the onset of a word that has undergone apophony is a stop, and where the resulting nucleus would normally be <vůe vůa>. The voiceless onset stops <p t k> are affected by a phenomenon called voiced labialization and voiced delabialization. When the voiceless onset stops <k> is precede <vůe vůa>, voicing occurs (with some exceptions) and it becomes labialised to <gvůe gvůa>, while <p t> become voiced, but labialization is lost, resulting in <bůe bůa> and <důe důa> respectively. Diachronically, this is due to earlier stages of Pine having post-aspirated initial stops which became unaspirated when labialised, and the subsequent loss of labialization for non-velar stops before front-rounded apophonised vowels. Characteristic of eastern dialects, the short labialised sequences [ᶣyœ̯ ᶣøœ̯ ʷuʊ̯] <vůe vůa vuo> are commonly lowered to [ᶣɛ ᶣæ ʷɔ] when preceded by a consonant (or even as an onset), for example ṛvůekki [ˈr̥ᶣyœ̯ʰkɪ] can be heard as [ˈr̥ᶣɛʰkɪ] (as if written ṛvekki). In eastern dialects where <ve> is [wɛ], this means that [wɛ] <ve> and [ɥɛ] <vůe> are two distinct sequences, which is not a distinction found in western dialects.

labialization and voicing

voicing with loss of labialization

§3.1.3.1.1 Dentolabialised Preaspiration

The front-rounded diphthongs <ůe ůa uo ey ay iů eu au iu> and <oa ỏ ủ> may become monophthongised and followed by /vʋ/ when followed by a preaspirated consonant. This is most commonly heard in colloquial, relaxed speech. Some speakers do not monophthongise diphthongs, especially in the eastern dialects, such that the result may be /vu̯ʋ/ or /vy̯ʋ/. The resulting cluster has initial consonant gemination as in most clusters, for example [ʋːp ʋːt ʋːk]. In eastern dialects with obvious rounding of the long /ɑː/ to [ɑː] <oa>, dentolabialization may also occur, as in toakkan [ˈtɑːʋːkan] ‘tall and thin (sem)’ instead of the more common [ˈtɑːʰkan]. This is also true of [oː] <ỏ> in eastern dialects, although this is not as common, for example in ỏttasṃi [ˈoːʋːtasːpm̥ɪ] ‘hips’ instead of the more common ỏttasṃi [ˈoːʰtasːpm̥ɪ].

For example, aukka ‘except’ may be pronounced [ˈɑʋːkɑ], [ˈɑu̯ʋːkɑ] or simply [ˈɑu̯ʰkɑ].

§3.1.3.1.2 Sibilant Fricative Preaspiration

The front closing diphthongs <ei ai ui ůi oi> and for some speakers <ỉ> may become monophthongised and are followed by /vɕ/ when followed by a preaspirated consonant. This is most commonly heard in colloquial, relaxed speech. Some speakers do not monophthongise diphthongs, especially in the eastern dialects, such that the result may be /vɪ̯ɕ/. The resulting cluster has initial consonant gemination as in most clusters, for example [ɕːp ɕːt ɕːk]. In some western dialects (mostly northwestern), compensatory lengthening occurs and [ɕ] is pronounced as [ʃ], such that gvaikkevyt ‘universe’ may instead be written as gvảṡkevyt, and ruikkįa as rủṡkįa.

§3.1.4 Degemination in Closed Syllables

A more recent phenomenon that was originally only found in certain western dialects but which has spread eastward amongst younger speakers is the reduction of geminate vowels to short tense vowels in closed syllables. This means that a word like ủbbukki [ˈuːpːʊʰkɪ] ‘midge’ becomes [ˈupːʊʰkɪ].

§3.1.5 Vowel Harmony

Some vowels in Pine cannot co-occur with other vowels inside the same word. This is called vowel harmony. At its core, the harmony system of Pine dictates that two groups of vowels are mutually exclusive. For clarity, <á> here will be used to represent /a/ in a stressed nucleus, and <a> will represent an unstressed /a/. The two mutually exclusive groups of vowels are called the opisthophonic vowels (u o á) and the prosthophonic vowels (y ů ả ai). Other vowels, namely <i e a> are amphiphonic.

The phonological requirement that front-rounded or prosthophonic vowels not be found in the same word as opisthophonic vowels results in something called serial harmony, where affixes may have two (bipartite serial harmony: front and back, or opisthophonic and amphiphonic) or three (tripartite serial harmony, opisthophonic, prosthophonic and amphiphonic) versions that follow the harmony alignment of the stem they are added to. Additionally, amphiphonic and opisthophonic series can undergo prosthothesis, which is the change of opisthophonic vowels into prosthophonic forms in cases where suffixes contain <u o> that must be harmonized to <y ů>. Therefore, amphiphonic and opisthophonic series may have additional prosthothesised forms. Additionally, the vowel <a> often has its own opisthophonic series, which in many cases is nearly identical to the standard opisthophonic series. Therefore, Pine suffixes (and in rare cases affixes) are organised into a number of series which harmonise with the stem onto which they are added.

Vowel harmony affects both prefixes and suffixes. In eastern dialects, unstressed <ai> is considered amphiphonic, but in standard language, it is replaced by <oi>. The nucleus of a root dictates the roundedness of the affixes, but serial harmony in nouns and adjectives often is dictated by the word-final vowel of the asematic form. When a front-rounded nucleus causes affixes to adapt to its front-roundedness, the change is called prosthothetic harmony, which dictates whether <ů> or <y> replaces <u o>, and <ả> replaces <oa>.

No matter the mechanism, Pine never allows opisthophonic vowels (<u o>) and prosthophonic vowels (<y ů>) to co-occur. In nouns for example, where we might expect a word-final -a to become -o, a front vowel in the root will cause it to become -ů, as in kaillįa ⟫ kảlįů 'rain, showers', where we would expect the change -a ⟫ -o, but prosthothesis requires the change -o ⟫ -ů.

§3.1.5.1 Prosthothesis

Prosthothesis is the phenomenon whereby the opisthophonic vowels <u o oa> become front-rounded to <y ů ả> or <ů ů ả> in affixes attached to stems that either contain a prosthophonic vowel, or contains an amphiphonic vowel but the affix itself requires prosthothesis.

Prosthothesis can affect affixes in one of two ways – strong prosthothesis changes <u> and <o> to <y> and <ů> respectively, whereas weak prosthothesis changes both <u> and <o> to <ů> unless the affix itself contains both <u> and <o>, in which case strong prosthothesis occurs. Each affix is either strongly or weakly prosthothesising, and nothing in the affix itself predicts which must occur. Strongly prosthothesising <u> is often written as <ụ> while weakly prosthothesising is assumed and not indicated. For example, -uƛ- and -yƛ-, the translative medials, can be represented as -ụƛ-.

If an affix is strongly prosthothesising, its prosthothetic form will always be listed separately. Weak prosthothesis is assumed in other cases. As such, if a suffix is given for example as -uṡot, then it is assumed to undergo prosthothesis to -yṡůt. For strongly prosthothesising suffixes like -uƛ- (translative suffix) or -ϕutuθ- (passive suffix), specific forms in -yƛ- and -ϕytyθ- must be given, as *-ůƛ- and *-ϕůtůθ- do not exist. The vast majority of thematic suffixes are weak, but many medials are strong. A few affixes have special derounded forms, such as the perfective -ung-, which has a prosthothesised forms in -ing-.

Below is the complete front-rounding harmony rule, where ∨ means ‘or’ and ∧ means ‘and’.

weak prosthothesis

strong prosthothesis

Below are examples contrasting weakly and strongly prosthothesising verbal affixes:

weak prosthothesis

strong prosthothesis

§3.1.5.2 Serial Vowel Harmony

Many affixes in Pine exist in multiple versions, called series, which are determined by vowel harmony. Opisthophonic series contain back vowels, amphiphonic series contain front vowels and a, and prosthophonic series contain front-rounded vowels. Opisthophonic and amphiphonic series also have prosthothesised versions, where <u o> are front-rounded to <y ů>. Suffixes can have tripartite or bipartite series. Bipartite series do not amphiphonic from prosthophonic (called front series), but an opisthophonic equivalent does exist. When the distinction is only between front or back as in bipartite series, the front vowel group is often called antopisthophonic. Affixes that do not exhibit any vowel harmony are not serial.

The illative plural suffix is an example of a tripartite suffix, having forms in -ittit (amphiphonic), -uttut (opisthophonic) and -yttyt (prosthophonic or prosthothesised opisthophonic). Case suffixes are tripartite, but most medials are bipartite, for example the causative medials -ed- (antopisthophonic) and -od- (opisthophonic). Suffixes that distinguish only opisthophonic and prosthophonic (grouping all non-front-rounded vowels together) are rare. The most notable example is the locutive medial -ỏm- (antiprosthophonic) and -eům- (prosthophonic), which derives verbs indicating manners or speech such as niṡkỏm- 'speak Pine'.

To illustrate this, the table below shows the seven versions of the inessive suffix -VmmVt: an opisthophonic form in -u, one in -a, an amphiphonic form in -i, a prosthophonic form in -y, and prosthothetic forms of the three opistho- and amphiphonic suffixes. It is not common for a suffix to have seven alternative forms, but the inessive singular is ubiquitous in Pine and exemplifies the phenomenon of serial harmony.

Below are more examples of verbal medials with prosthophonic and antiprosthophonic serial harmony:

§3.1.6 Syllable Structure & Vowels

Pine vowels are affected by syllable structure, both in their pronunciation (short or long lax quality) and through rhizal apophony. Pine distinguishes three types of syllable closure: open, closed and lytic syllables. Lytic syllables are semi-open: they are functionally open for diphthongised vowels and in the case of <u o> with regard to rhizal apophony. For these two vowels, the output of rhizal apophony differs in lytic syllables compared to either open or closed syllables. Geminate vowels and true diphthongs are not affected by syllable structure. Syllable closure is determined by coda complexity:

Geminate sonorant-fricatives in Pine are:

[ʋː mː m̥ː nː n̥ː ɲː ɲ̊ː ŋː hː ɕː jː ðː sː ʃː xː ɣː rː r̥ː lː ɬː ʎː ʎ̥ː]

<vv mm ṃṃ nn ṇṇ nnį ṇṇį hh hhį įį ḍḍ ss ṡṡ tġ ġġ rr ṛṛ ll ḷḷ llį ḷḷį>

§3.2 Stress

Pine has immobile stress fixed on the initial syllable of a root word, called archeostatic stress. Virtually every word in Pine retains its archeostatic stress no matter what derivational affixes it may gain. Prefixes do not change which syllable is stressed. Because stress is not explicitly indicated by Pine orthography and prefixes are extremely common, it may be difficult to know which syllable is the root of a word. Secondary stress occurs on the penultimate vowel, which if lax and before a non-geminated simple consonant, is usually lengthened. Every second syllable after the stressed syllable of a word may be slightly stressed. Only in cases where there are more than two syllables before a stressed vowel will there be lengthening of lax vowels before simple consonants in prefixes. This is called peninitial lengthening.

Consider the various forms of the verb alk- ‘give, allow’, where the initial a- bears the primary stress of the word.

§3.3 Consonants

Pine has a rather large consonant inventory of 36 consonants and three allophones, which contains a few uncommon sounds, such as five different lateral consonants, voiceless nasals, a glottal stop, preaspirated, preglottalised as well as labialised consonants. It also has a large number of consonant clusters typically found in the stressed coda, a smaller but significant number of onset clusters, but only a small number of consonants allowed word-finally.

Below is a table summarizing the consonants found in Pine, both phonemes (//) and allophones ([]).

Pine contains a series of preaspirated consonants, /ʰp ʰt ʰc ʰk ʰts ʰt͡ʃ/ <pp tt kkį kk tts ttṡ> as well as preglottalised consonants found with homorganic stops /ʔpm ʔtn ʔtɲ ʔkŋ/ <bm dn dnį kn> or without homorganic stops /ʔw ʔj ʔr/ <ḥv ḥį ḥr>. Note that /t͡s t͡ʃ/ <ts tṡ> may be either preaspirated or geminated when long, such that <tts ttṡ> represents either [t͡sː t͡ʃː] or [ʰt͡s ʰt͡ʃ]. Preaspiration of <tts ttṡ> is most common in eastern dialects. All consonants except /p c ɟ f θ s ʃ ɕ j ʎ ʎ̥/ are found with labialization: /mʷ m̥ʷ nʷ n̥ʷ ŋʷ ŋ̊ʷ ɲʷ ɲ̊ʷ bʷ tʷ dʷ kʷ gʷ ʔʷ ðʷ xʷ ɣʷ t͡sʷ t͡ʃʷ rʷ r̥ʷ lʷ ɬʷ t͡ɬʷ/, and in some dialects /ʃʷ sʷ/ are also found . These can be front-rounded (represented by [Cᶣ]. The consonants /ɥ θ/ are allophones or /w ð/, occurring only in certain environments. The uvular stop /q~χ/ appears as an allophone of /x/ <tġ>, more commonly amongst male speakers in the west, although it has become more common amongst younger speakers in the east.

§3.3.1 Nasals

Pine has four pairs of contrasting voiced and voiceless nasals: [m m̥] <m ṃ>, [n n̥] <n ṇ>, [ɲ ɲ̊] <nį ṇį> and [ŋː ŋ̊ː] <ng>, which are always geminated intervocalically. Voiceless nasals are always geminated intervocalically, unless root initially preceded by prefixes. The other nasals may also appear as geminates intervocalically: [mː m̥ː] <mm ṃṃ>, [nː n̥ː] <nn ṇṇ> and [ɲː ɲ̊ː] <nnį ṇṇį>. The velar nasal /ŋ/ only appears voiceless as [ŋ̊] allophonically in the clusters [ŋ̊ːk] <nk>, [ɬː(k)ŋ̊] <lkn>, [θː(k)ŋ̊] <ḍkn>, [r̥ː(k)ŋ̊] <rkn>, [t͡ɬː(k)ŋ̊] <ƛkn> and [kːŋ̊] or [kːn̥] <kṇ> . All nasals can be preglottalised, in which case they appear with a homorganic stop: [ʔpm] <bm>, [ʔtn] <dn>, [ʔtɲ] <dnį> and [ʔkŋ] <kn>. There is also a corresponding homorganic voiceless series: [pːm̥] <pṃ>, [tːn̥] <tṇ>, [tːɲ̊] <tṇį>, [kːŋ̊] <kṇ>, although <kṇ> can also be pronounced [kːn̥]. These voiceless homorganic nasal clusters are also found with preaspiration (in which case the initial consonant is not geminated): [ʰpm̥ ʰtn̥ ʰtɲ̊ ʰkŋ̊~ʰkn̥] <pṃ tṇ tṇį kṇ> in western dialects. The clusters [tm̥] <tṃ> or [pm̥] <pṃ> and [tn̥ tɲ̊] <tṇ tṇį> can appear word-initially, where <pṃ> is more common outside of eastern dialects, as well as in the initial clusters [spm̥] <sṃ> and [stn̥ stɲ̊] <sṇ sṇį>. A handful of words have word-initial [km̥] <kṃ> and [kn̥ kɲ̊] <kṇ kṇį>, for example kṃoa [km̥ɑː] ‘species’ and kṇįuol [kɲ̊ʊu̯ɬ] ‘small hammer’.

The combination of a nasal and <v> is not allowed except for [ŋːʷ~ŋːᶣ ŋ̊ːkʷ~ŋ̊ːkᶣ ŋːkʷ~ŋːkʷ] <ngv nkv nkkv>. The nasals <m ṃ n ṇ ṇį ng> can be labialised, although only rarely, in which case /ʷ/ is represented by <ů> or <u>. Only the following sounds are found after non-palatalised nasals word-initially: [ɥɛɪ̯ ɥɛy̯ ɥæɪ̯ ɥæy̯ wɑu̯] <ůei, ůey, ůai, ůay, uoa>.

§3.3.2 Plosives

Pine has four pairs of plosives, or stops, that contrast voicing: labial plosives [p b] <p b>, dental plosives [t d] <t d>, palatal plosives [c ɟ] <kį gį> and velar plosives [k g] <k g>. In addition, there is also a glottal stop [ʔ] <ḥ>, which can occur word initially. Palatal plosives can be variously pronounced [c ɟ] or [kj gj]. Many speakers have /k g/ as /c ɟ/ before front vowels. Palatal plosives only occur before front vowels in a handful of words (kįỉt [ˈciːt] or [kjiːt] ‘reindeer calf in its first year’) in the stressed syllable, and are otherwise not present before front vowels in unstressed syllables, for example in ṛuikkįa [ˈr̥ʊ̯ɪʰcɑ] ‘comb’, which has the sematic form ṛveiggi [ˈr̥ᶣɛɪ̯kːɪ] instead of *ṛveiggįi. Stops do not have postaspiration. Intervocalically, long voiceless stops are preaspirated: [ʰp ʰt ʰk ʰc] <pp tt kk kkį>, but the geminated glottal stop is most commonly only geminated [ʔː] <ḥḥ>, although [ʰʔ] also exists. This is also true of labialised intervocalic voiceless stops, but the sequences *<pv kįv gįv> are not found. Intervocalically, labialised voiceless stops are preaspirated [ʰtʷ~ʰtᶣ ʰkʷ~ʰkᶣ] <tv kv>, but the sequence <pv> is not allowed and <ḥv> is [ʔʷː~ʔᶣː]. Word-initially, [bʷ dʷ] <bv dv> only exist as voiced and before back vowels, but velar stops distinguish voicing: [kʷ~kᶣ gʷ~gᶣ] <kv gv>. Intervocalically, voiced labialised stops are geminated and devoiced to [tʷː~tᶣː kʷː~kᶣː] <dv gv>, with <bv> not occurring outside of the onset of a root and only before back vowels. The sequences <bve bvi> only occurs in onomatopoeiae, and <dv> only occurring before front vowels as [dᶣ] outside of stressed syllables. The rather uncommon sequence <ḥḥį> can be heard as [ʔːɕ] or as [ʔːj], depending on the speaker. There is some variation regarding preaspiration, where some speakers replace it with preglottalization, such that <pp tt kk> may be pronounced [ʔp ʔt ʔk]. This is common in northeastern dialects, and used to be common in northwestern dialects as well. Intervocalic non-geminated voiceless stops are less common but do exist. Voiced geminate stops become voiceless: [pː tː kː cː] <bb dd gg ggį>. Voiced stops occur both word-initially and word-internally: [b d g ɟ] <b d g gį>. Only three stops are allowed word-finally: [t, d~ð~ʔ~j, ː] <t>, [k] <k> and [kʷ] <ḳ>.

Word-finally, <t> has a range of possible pronunciations. The standardised pronunciation is [t] in all cases except before a following vowel (and commonly v-, especially va/vu 'and'), in which case it is pronounced [h]. When hyphenated <-t>, it takes on the pronunciation of any following consonant or glide without causing preaspiration of stops, such that <-t b> and <-t p> are pronounced the same. This is indicated by [ ͜ ], for example a gami-t da [ɑ ˈgɑːmɪd ͜ dɑ] ‘as for me, I’ . The sequence [d ͜ d] is pronounced as [tː], and a įoa-t ka [ɑ ˈjɑːk ͜ kɑ] ‘right now’, where [k ͜ k] is pronounced [kː]. In normal, relaxed or colloquial registers, word-final <t> is pronounced as its hyphenated form before consonants, causing gemination of the following consonant: nảt įeįįa nỉ [næːj ͜ ˈjɛjːɑ niː] 'they are here'. Hyphenation of <-t> to a final syllable does not close it, which means that <a> appears as [ɑ] even before a geminated consonant.

If a hyphenated <-t> is required due to a clitic on a word that ends in <t>, the hyphenated <-t> replaces it, as in Nįiḷḷut [ˈɲɪɬːʊt] but a Nįiḷḷu-t da [ɑ ˈɲɪɬːʊd ͜ dɑ].

One important exception to the standard pronunciation of word-final <t> is that when a hyphenated <-t> is added onto an already hyphenated clitic, as in mi na-t de, the hyphen is dropped. This does not affect the pronunciation of the hyphenated <-t>, which can be tricky for learners as it renders the distinction between word-final <t> and <-t> opaque in those speakers who have a geminating final <-t> pronunciation.

Compare the standard (s) and more colloquial (c) pronunciation of word-final <t> in the following example.

sġauddot giḷḷittut simindat na tsiskvu-t ke įi?

shf-aut-tu-t giLLittut simint-at na c-iskfu_0 ki- ji?

(s) [ˈsxɑu̯tːɔt ˈgɪɬːɪʰtʊt sɪˈmɪnːdat nɑ ˈt͡sɪsːkʷʊːk ͜ kɛ jɪ]

(c) [ˈsxɑu̯tːɔg ͜ ˈgɪɬːɪʰtʊs ͜ sɪˈmɪnːdan ͜ nɑ ˈt͡sɪsːkʷʊːk ͜ kɛ jɪ]

si=hauddot giḷḷi-ttut si-mind-at na ts-isk-vu ke–t įi

tomorrow.morning earl.ier cis-come-dep interr 2.ag-can-loc.dub interr interr

‘could you could earlier tomorrow morning?’

One of the distinguishing features of western dialects is the loss of preaspiration of stops, resulting in an open syllable but a short vowel, for example ittava ‘thank you’ is famously pronounced [ˈɪtɑːwɑ] in the west instead of the more standard [ˈɪʰtɑːwɑ].

§3.3.3 Fricatives

Excluding the two semivowels /w ɥ/ and /j/, Pine has 10 fricatives: /f ʋ θ ð s ʃ ɕ x ɣ h/. Only /ð x ɣ/ can be labialised, of which only [xʷ~xᶣ ɣʷ~ɣᶣ] can occur word-initially. When preceded by a voiceless consonant in clusters, <ġv> is allophonically realised as [-xʷ~-xᶣ]. Some speakers voice [f] to [v] in all positions. The dental fricative /ð/ <ḍ> has both a voiced [ð] and voiceless [θ] realization, but [θ] is only present in a handful of clusters, namely [θːk] <ḍk>, [θː(k)ŋ̊] <ḍkn>, [θːpm̥] <ḍm>, [θːtn̥] <ḍn> and [θːj] <ḍḍį>, although the last two voiced after non-short vowels: [ðːm] and [ðːn] respectively. It only appears intervocalically and in clusters, but the intervocalic non-geminated [ð] <ḍ> is no longer found in any dialect – speakers pronounce it as [h] <h>, however <ḍį> [ðj~ðːj] is never found as <hį>, as in ỉḍįon ‘spring (sem)’. Not all dialects distinguish <ḍḍį> and <ḍį>, but those that do usually have <ḍḍį> as [θːj] or [θːɕ] and <ḍį> as [ðːj], which is the standardised pronunciation. The dental fricative pronunciation is conserved in the verbs miuḍ- ‘breathe’ and miḍ- ‘want, like’ only when it is geminated by a v-versioner, for example: gamiuḍḍa ‘I breathe’ (<*ga-miuḍ-va) but tṃiuhan ‘s/he breathes’ (<*t-miuḍ-an) and gamiḍḍa (<*ga-miḍ-va) but gamihet ‘I like it, I want it’ (*<ga-miḍ-et).

The fricatives /s ʃ/ <s ṡ> are the only fricatives that can appear in word-final position in Pine, and can be found as geminates intervocalically: [sː ʃː] <ss ṡṡ>. Word-initially, [stn̥ spm̥] are written <sṇ sṃ> but word internally, [sːtn̥ sːpm̥] are written <stṇ spṃ>. The fricative /ɕ/ <hį> can appear word-initially, but intervocalically it is always geminated to [ɕː] <hhį>.

Pine has two velar consonants, /x ɣ/ <tġ ġ>. The latter is allophonically realised either as a velar [x] or an uvular [χ] (although only [x] is used here) when directly adjacent to a voiceless consonant, such as in the clusters [xːt] <ġt>, [sːx] <sġ> [t͡sːx] <tsġ>, [r̥ːx] <rġ>, [ɬːx] <lġ> and [xːt͡ɬ] <ġƛ>. The voiceless [x] <tġ> is always geminated intervocally and can be labialised as [xʷː xᶣː] <tġv> word-initially or intervocally. The voiced version <ġ> is pronounced [ɣ] <ġ> word-initially, intervocally and in most clusters, for example [rːɣ] <rġġ>, [lːɣ] <lġġ>. It also has labialised forms in [ɣʷː ɣᶣː] word-initially and intervocally, and in some clusters. In three clusters, it alternates between voiceless and voiced forms depending on its position in the word or the preceding vowel: [ðːɣ] <ḍġ>, [ɣːn] <ġn> and <ġm> [ɣːm] are found in unstressed syllables and following a long or geminate vowel or diphthong in the root nucleus. They are found as [θːx] <ḍġ>, [xːtn̥] <ġn> and [xːpm̥] <ġm> after after a short root nucleus. The voiced velar /ɣ/ can be geminated to [ɣː] <ġġ> intervocally, and in clusters, <ġġ> indicates voicedness while <ġ> voicelessness. Some speakers contrast a short and long form, in which case <tġ> represents the short form and <tġġ> represents the geminated form. Some speakers across all dialects replace the phoneme [x] by [χ] or [q], including when labialised, then as [χʷ~χᶣ] or [qʷ~qᶣ].

The phoneme /h/ can be found word-initially or intervocalically as [h], or geminated intervocalically to [hː] <hh>, sometimes realised as [ɦː], especially in northern dialects. The word-initial sequence /h+v/ is realised as [p] <p>, for example when a third-person inanimate unagentive (h-) combines with a v- initial verb: h-vatṇuƛami ⟫ patṇuƛami [ˈpatːn̥ʊt͡ɬɑːmɪ] ‘it closes’. Tuġvut uses a special character to distinguish the normal <p> from a <p> historically resulting from /h+v/.

The cluster <ṡv> [ʃʷ~ʃᶣ] is found in standard language only in a few instances, the most common being iṡvi ‘parent’ and piṡvi ‘lynx’, although many have <ṡṡ> [ʃː] instead (iṡṡi, piṡṡi). In contrast, <sv> is not attested except in some western dialects in the case of a third-person animate unagentive prefix s- before a verb that begins with v-, for example svassan ‘s/he is ashamed’ instead of the much more common sassan.

§3.3.4 Affricates

Pine only has two affricates, /t͡s/ <ts> and /t͡ʃ/ <tṡ>. Both can be found word-initially, intervocalically or in clusters and -ts occurs word finally either by itself or in the cluster [-n̥t͡s] <-nts>, the only true word-final cluster in Pine, though nearly all speakers reduce this to -s in non-formal registers. When geminated, they can be preaspirated or simply geminated: [t͡sː~ʰt͡s] <tts> and [t͡ʃː~ʰt͡ʃ] <ttṡ>. Gemination is used in this book as it is considered more standard. Some speakers of eastern dialects tend to voice them to [d͡z] and [d͡ʒ] intervocalically. Both /t͡s/ and /t͡ʃ/ can be labialised word-initially as [t͡sʷ~t͡sᶣ t͡ʃʷ~t͡ʃᶣ], but they are more commonly found with gemination instead of labialization in standardised language.

§3.3.5 Trills

Pine only two voice contrasting trills: /r/ <r> and /r̥/ <ṛ>, which can appear word-initially, word-internally and <-r> can appear word-finally, in which case it is devoiced to [r̥]. Both can be geminated intervocalically to [rː] <rr> and [r̥ː] <ṛṛ>. Many clusters have voice and voiceless pairs, where nasals also have a homorganic stop when voiceless, although /ŋ/ only has a voiceless version : [rːm] <rm> and [r̥ːpm̥] <rpṃ>, [rːn] <rn> and [r̥ːtn̥] <rtṇ>, [rːpn] <rbn> and [r̥ːpn̥] <rpṇ>, [rːkŋ] <rgn> and [r̥ːkŋ̊] <rkn>, [rːt͡ɬ] <rƛƛ> and [r̥ːt͡ɬ~r̥ːɬ] <rƛ>, [rːp] <rpp> and [r̥ːp] <rp>, [rːt] <rtt> and [r̥ːt] <rt>, [rːt͡s] <rtts> and [r̥ːt͡s] <rts>, [rːt͡ʃ] <rttṡ> and [r̥ːt͡ʃ] <rtṡ>, [rːk] <rkk> and [r̥ːk] <rk>, [rːɣ] <rġġ> and [r̥ːx] <rġ>, as well as [rːj] <rį> and [r̥ːɕ~r̥ːj~r̥ːt͡ʃ] <ṛį>. In addition, the voiced stops <d g> also form clusters with /r/: [rːd] <rd> and [rːg] <rg>. Word-initial /r r̥/ can also be labialised: [rʷ~rᶣ] <rv-> and [r̥ʷ~r̥ᶣ] <ṛv->, but the /r/ is geminated word-internally and <ṛv> is always [r̥ːf]. There are both voiced and voiceless labialised versions of <r-> clusters contain velars or uvulars, for example [rːɣʷ~rːɣᶣ] <rġġv> and [r̥ːxʷ~r̥ːxᶣ] <rġv>. There exist a few words that contain the cluster [ʔr] <ḥr> (sometimes written <kr>), for example raḥri [ˈraʔrɪ] ‘raven’.

§3.3.6 Laterals

Pine contains five different laterals: /l ɬ t͡ɬ ʎ ʎ̥/ <l ḷ ƛ lį ḷį>. All can be geminated, but /t͡ɬː/ <ƛƛ> may also be found as [ɮː] or [d͡ɮː] intervocalically in eastern dialects, although many speakers instead preaspirate it to [ʰt͡ɬ] in the same environment. Standard language has true geminates: [lː ɬː ʎː ʎ̥ː t͡ɬː] <ll ḷḷ llį ḷḷį ƛƛ>. The lateral <l> can appear word-finally, where it is devoiced to [ɬ]. Like the nasals <m n>, <l ḷ lį ḷį> cannot be found word initially directly preceding <v>, but <ƛ> can, however <l ḷ> can be followed by <vei vai vey vay vau> in which case the they are written <lůei lůai lůey lůay luoa> and <ḷůei ḷůai ḷůey ḷůay ḷuoa> respectively. Intervocalically, <ḷv> is always [ɬːf]. The lateral /l ɬ/ have most of the same voice contrasting consonant clusters as /r r̥/ , but unlike /r/, the cluster [lːŋ] <lgn> is allowed, and /l ɬ/ cannot cluster with /t͡ɬ/ : [lːm] <lm> and [ɬːpm̥] <lpṃ>, [lːn] <ln> and [ɬːtn̥] <ltṇ>, [lːpn] <lbn> and [ɬːpn̥] <lpṇ>, [lːŋ] <lgn> and [ɬː(k)ŋ̊] <lkn>, [lːp] <lpp> and [ɬːp] <lp>, [lːt] <ltt> and [ɬːt] <lt>, [lːt͡s] <ltts> and [ɬːt͡s] <lts>, [lːt͡ʃ] <lttṡ> and [ɬːt͡ʃ] <ltṡ>, [lːk] <lkk> and [ɬːk] <lk>, [lːɣ] <lġġ> and [ɬːx] <lġ> as well as the voiced clusters [lːb] <lb>, [lːd] <ld> and [lːg] <lg>. As with <r-> clusters, <l-> clusters containing velars or uvulars can occur as voiced or voiceless labialised clusters, for example [lːkʷ~lːkᶣ] <lkkv>, [ɬːkʷ~ɬːkᶣ] <lkv> and [lːgʷ~lːgᶣ] <lgv>. Note that the letter <ƛ> is either capitalised as <Ł> or <Ƛ> (in which case <ƛƛ> becomes <ꟜꟜ>), such that the endonym of the Pine language, Nisƛikketṡin, when capitalised is NIṠꟜIKKETṠIN. The letter <Ł> is most common in print, as <Ƛ> is not widely available, but it is the standard in handwriting. This book uses <Ƛ>.

§3.3.7 Consonant Clusters

Pine has a large number of possible consonant clusters, the majority of which appear in the stressed coda of roots. Clusters can up to three consonant phonemes, for example <rtsƛ> and <kts>, as in urtsƛis [ˈʊr̥ːt͡st͡ɬɪs] ‘itchy bite’ or raktsi [ˈrakːt͡sɪ] ‘skin bladder’, or three consonants and <v>, as in gatṡẻlġġvet [gɑˈt͡ʃeːlːɣɥɛt] ‘I cook it’ . A smaller but significant number of clusters appear word-initially, and only one, [-n̥t͡s] <-nts> can appear word-finally (though nearly universally replaced by -s in non-formal registers).

All consonant clusters have a geminated initial element, and all clusters ending in <t k d g ts tṡ ġ ƛ> can be followed by <v>, but clusters ending in <pp tt kk>, can only be followed by <v> once they have been voiced, such that *<lpp-v ltt-v lkk-v> become <lb ldv lgv> but <lb lttv lkkv> also exist. Clusters whose last element is <tts ttṡ ġġ ƛƛ> can be followed by <v>, but many speakers either drop the <v> or more rarely devoice the cluster in such cases, such that gatṡẻlġġvet [gɑˈt͡ʃeːlːɣᶣɛt] ‘I cook it’ can also be heard as gatṡẻlġġet [gɑˈt͡ʃeːlːɣɛt] or gatṡẻlġvet [gɑˈt͡ʃeːɬːxᶣɛt]. The dropping of <v> after these double consonants is becoming the norm. Note that the sequences *<lb-v rb-v> are realised as [lːp rːp] <lpp rpp>.

§3.3.7.1 Word-Initial Clusters

Pine allows a number of word-initial clusters, or about 31, although many only appear in a handful of roots, and many are the result of prefixes and do not occur otherwise or only very rarely. The table below excludes simple labialised consonants but includes clusters that contrast labialization.

§3.3.7.2 Word-Internal Clusters

All word-initial clusters are also allowed word-internally, and a significant number of clusters are only allowed word-internally. Unlike word-initial clusters, all clusters found word-internally have a geminated first element.

§3.3.7.2.1 Labial-Initial Clusters

Labial-initial clusters have a labial consonant as their first element, either /m m̥/, /n n̥/ or /ŋ ŋ̊/. A handful of words contain the cluster [mːɣ] <mġġ>, namely ỉmġġot ‘after which’, ủmġġot ‘after which (not)’, ỉmġġit ‘without which’ (more commonly found as ỉmmen uġġit or ỉmmuġġit) and ủmġġit (more commonly found as ủmmen uġġit or ủmmuġġit).

§3.3.7.2.2 Plosive-Initial Clusters

Plosive-initial clusters have a plosive as their initial element, either /p/, /t/ or /k/, including preaspirated stops.

Non-homorganic clusters with an initial plosive and a final nasal may be preaspirated, especially in eastern dialects, and in some cases the plosive may be dropped all together leaving a preaspirated voiceless nasal.

Many speakers realised [tːpm̥] <tṃ> as [ʰpm̥] or even [ʰm̥], and <pṇ> can likewise be preaspirated as [ʰpn̥] or [ʰn̥]. Similarly, <kṃ> can be realised as [kːpm̥], with preaspiration as [ʰkm̥] or as [ŋ̊m̥~ʰŋ̊m̥], especially in western dialects. The cluster <kṇ> is pronounced [kːŋ̊] in standard registers but it may be also be preaspirated as [ʰkŋ̊] or [ʰtn̥]. Likewise, <kl> is either [kːɬ] or preaspirated to [ʰkɬ] or even [ʰt͡ɬ]. In western dialects, <kl> is often realised as [ʔɬ].

§3.3.7.2.3 Fricative-Initial Clusters

Fricative-initial clusters have a fricative as their initial element: /ð ɣ s ʃ/. Note that the clusters <ḍm ḍn ḍnį> can be pronounced as voiced [ðːm ðːn ðːɲ] or voiceless [θːpm̥ θːtn̥ θːtɲ̊], as can <ġm ġn> [xːpm̥~ɣːm xːtn̥~ɣːn]. They are generally voiced after a long vowel but voiceless after a short vowel in standard language, but there some speakers only have one or the other. In unstressed syllables, they are always voiced. For example, siġma ‘rainy weather’ is [ˈsɪxːpm̥ɑ] but ƛiggaġmogi ‘directions’ is [ˈt͡ɬɪkːaɣːmɔːgɪ]. The cluster <ḍkn> can be with a homorganic stop, then [θːkŋ̊] or without as [θːŋ̊]. The cluster <ḍġ> is quite rare and can be found voiced [ðːɣ] and voiceless [θːx], both of which are in free variation, but typically standard language favours the voiceless version.

§3.3.7.2.4 Affricate-Initial Clusters

Affricate-initial clusters have an affricate as their initial element, either /t͡s/ or /t͡ʃ/. Some speakers pronounce all affricate-initial clusters by dropping the [t] and replacing it with [ʔ], such that <tsṃ> is pronounced [ʔspm̥] or [ʔsm]. This is not a dialect specific feature but rather associated with sociolects, especially with masculine or ‘rough’ language. The preglottalised pronunciation of affricate-initial and stop-final clusters has in recent times however become much more common, such that [ʔst ʔsk ʔʃt ʔʃk] <tst tsk tṡt tṡk> is no longer as strongly associated with masculine registers and has become a normal, non-stigmatised realization of these clusters.

The cluster <tsƛ> is often reduced to [t͡sːɬ].

§3.3.7.2.5 Trill-Initial Clusters

Trill-initial clusters have a trill as their initial element, either /r/ or /r̥/. The rare cluster [r̥ːt͡st͡ɬ] <rtsƛ> is sometimes reduced to [r̥ːst͡ɬ] <rsƛ> or simply [t͡sːt͡ɬ] <tsƛ>. Not all speakers have <rbn rpṇ> and may instead use <rn rtṇ> or <bn pṇ>, which is only found in two roots: ƛarbna ‘bladderwhack’ and related terms, and dirbna ‘European cisco’.

§3.3.7.2.6 Lateral-Initial Clusters

Lateral-initial clusters have an lateral as their initial element: /l ɬ t͡ɬ/. The rare clusters [lːɲ ɬːtn̥] <lnį ltṇį> are often realised as [ʎːɲ ʎ̥ːtɲ̊]. Only a handful of roots have either: kvilnįu ‘sinew thread’ and related terms, ůlnįůn ‘back of knee/elbow’, and an alternative form of tṇįultṇuṡ ‘covered in bites’ is tṇįultṇįuṡ. The clusters <lttv lkkv> are often written and pronounced <ldv lgv> instead in less formal registers.