Skip to Content

Night Speech Phonology

https://www.jamescolton.com/worldbuilding/night-speech/phonology/

Consonants

Labiodental Dental Alveolar Postalveolar Velar Uvular
Nasal /ɱ/ /n/
Stop /t d/ /k/
Fricative /v/ /θ ð/ /s z/ /ʃ ʒ/ /x/
Approximant /χ ʁ/
Lateral Approximant /l/
Consonants
IPA Pronunciation
/ɱ/ English ⟨m⟩ with upper teeth against lower lip, as in ⟨circumvent⟩
/n/ English ⟨n⟩ as in ⟨night⟩
/t/ English ⟨t⟩ as in ⟨time⟩
/d/ English ⟨d⟩ as in ⟨dark⟩
/k/ English ⟨k⟩ as in ⟨kill⟩, never ⟨c⟩ as in ⟨cell⟩
/v/ English ⟨v⟩ as in ⟨vapor⟩
/θ/ English ⟨th⟩ as in ⟨thin⟩
/ð/ English ⟨th⟩ as in ⟨then⟩
/s/ English ⟨s⟩ as in ⟨sigh⟩
/z/ English ⟨z⟩ as in ⟨zero⟩
/ʃ/ English ⟨sh⟩ as in ⟨shy⟩
/ʒ/ English ⟨s⟩ as in ⟨vision⟩
/x/ Scottish ⟨ch⟩ as in ⟨loch⟩
/χ/ Guttural hiss, with optional growl or purr
/ʁ/ French ⟨r⟩ as in ⟨notre⟩, with optional growl or purr
/l/ English ⟨l⟩ as in ⟨lake⟩, never as in ⟨cool⟩
Consonant Pronunciation Guide

Night Speech has a late voicing onset time. Prolonged voiced consonants may even be pronounced as an unvoiced phoneme followed by a voiced one (e.g., /zː/ may be pronounced as [sz]).

[ɱ̥ ɱ], [n̥ n], and [l̥ l] are complementary allophones of /ɱ/, /n/, and /l/, respectively. The voiced phonemes in each of those pairs only occur adjacent to other voiced consonants.

Although [χ ʁ] are technically fricatives according to the IPA, in Night Speech they may be formed as either fricatives or approximants. For the purposes of phonotactical constraints, they are always considered approximants.

Doubled nasals, fricatives, and approximants are simply prolonged. The second in a doubled stop is fricativized (e.g., /kk/ is pronounced [kx]). Note that alveolar /t d/ fricativize into dental /θ ð/ (e.g., /tt/ is pronounced [tθ]).

Vowels

Front Central Back
Close /i/ /ɯ/
Close-Mid /ɪ/ /ɤ/
Open-Mid /ɛ/ /ʌ/
Open /ä/
Vowels
IPA Pronunciation
/i/ English ⟨ee⟩ as in ⟨free⟩
/ɯ/ English ⟨u⟩ as in ⟨lure⟩, but without rounding the lips
/ɪ/ English ⟨i⟩ as in ⟨bit⟩
/ɤ/ English ⟨oo⟩ as in ⟨foot⟩
/ɛ/ English ⟨e⟩ as in ⟨bed⟩
/ʌ/ Something between English ⟨u⟩ as in ⟨gut⟩ and ⟨o⟩ as in ⟨not⟩
/ä/ English ⟨a⟩ as in ⟨father⟩
Vowel Pronunciation Guide

Although seven vowels are given above, in practice they can often be reduced to just three: /ɪ ɤ ä/. All the others are allophones in most cases:

  • [i ɪ ɛ] are allophones of /ɪ/
  • [ɯ ɤ ʌ] are allophones of /ɤ/

Note that, in contrast to English, the lips are never rounded when pronouncing back vowels.

Doubled or long vowels are pronounced as two short vowels separated by a non-syllabic /χ/ (e.g., /äː/ is pronounced [äχä]).

When /ɛ/ appears at the end of a word following a consonant, it tends to centralize toward the schwa sound [ə].

Diphthongs

IPA Pronunciation
/ɪ̯ɛ/ English ⟨ye⟩ as in ⟨yet⟩
/ɤ̯ɪ/ English ⟨wi⟩ as in ⟨win⟩
/ɤ̯ä/ English ⟨wa⟩ as in ⟨water⟩
/ɛɪ̯/ English ⟨ay⟩ as in ⟨day⟩
/äɪ̯/ English ⟨i⟩ as in ⟨spine⟩
Diphthong Pronunciation Guide

Phonotactics

Night Speech syllables adhere to the following rules:

  • Syllables follow the pattern (C)(C)V(C)(C), that is:
    1. Up to two optional consonants
    2. A required vowel (which may be short, long, or a diphthong)
    3. Up to two optional consonant
  • Onset consonant clusters adhere to the following order:
    1. Nasal or stop
    2. Fricative
    3. Approximant
  • Coda consonant clusters are only allowed at the end of a word, and adhere to the opposite order:
    1. Approximant
    2. Fricative
    3. Nasal or stop
  • Consonant pairings from the same level of the hierarchy are only allowed between onset fricatives, and only if they adhere to the following order:
    1. Dental
    2. Alveolar
    3. Labiodental
  • Postalveolar fricatives may not be paired with other fricatives
  • Adjacent consonants must agree in voicing
  • /x/ may only be paired with /k/ and approximants, or /t/ in codas

For ease of reference, here are the romanizations of all the possible Night Speech onsets:

⟨c⟩ ⟨cch⟩ ⟨cl⟩ ⟨cr⟩ ⟨cth⟩ ⟨cv⟩ ⟨cz⟩ ⟨czh⟩
⟨ch⟩ ⟨chl⟩ ⟨chr⟩
⟨d⟩ ⟨ddh⟩ ⟨dl⟩ ⟨dr⟩ ⟨dv⟩ ⟨dz⟩ ⟨dzh⟩
⟨dh⟩ ⟨dhl⟩ ⟨dhr⟩ ⟨dhv⟩ ⟨dhz⟩
⟨h⟩
⟨l⟩
⟨m⟩ ⟨mdh⟩ ⟨ml⟩ ⟨mr⟩ ⟨ms⟩ ⟨msh⟩ ⟨mth⟩ ⟨mv⟩ ⟨mz⟩ ⟨mzh⟩
⟨n⟩ ⟨ndh⟩ ⟨nl⟩ ⟨nr⟩ ⟨ns⟩ ⟨nsh⟩ ⟨nth⟩ ⟨nv⟩ ⟨nz⟩ ⟨nzh⟩
⟨r⟩
⟨s⟩ ⟨sl⟩ ⟨sr⟩ ⟨sv⟩
⟨sh⟩ ⟨shl⟩ ⟨shr⟩
⟨t⟩ ⟨tl⟩ ⟨tr⟩ ⟨tth⟩ ⟨tv⟩ ⟨tz⟩ ⟨tzh⟩
⟨th⟩ ⟨thl⟩ ⟨thr⟩ ⟨thv⟩ ⟨thz⟩
⟨v⟩ ⟨vl⟩ ⟨vr⟩
⟨z⟩ ⟨zl⟩ ⟨zr⟩ ⟨zv⟩
⟨zh⟩ ⟨zhl⟩ ⟨zhr⟩

Onsets

And all the possible Night Speech codas:

⟨c⟩ ⟨ct⟩
⟨ch⟩ ⟨chc⟩
⟨d⟩
⟨dh⟩ ⟨dhd⟩
⟨l⟩ ⟨lc⟩ ⟨lch⟩ ⟨ld⟩ ⟨ldh⟩ ⟨lm⟩ ⟨ln⟩ ⟨ls⟩ ⟨lsh⟩ ⟨lt⟩ ⟨lth⟩ ⟨lv⟩ ⟨lz⟩ ⟨lzh⟩
⟨m⟩
⟨n⟩
⟨r⟩ ⟨rc⟩ ⟨rch⟩ ⟨rd⟩ ⟨rdh⟩ ⟨rm⟩ ⟨rn⟩ ⟨rs⟩ ⟨rsh⟩ ⟨rt⟩ ⟨rth⟩ ⟨rv⟩ ⟨rz⟩ ⟨rzh⟩
⟨s⟩
⟨sh⟩
⟨t⟩
⟨th⟩ ⟨thc⟩ ⟨tht⟩
⟨v⟩ ⟨vc⟩ ⟨vd⟩ ⟨vt⟩
⟨z⟩ ⟨zc⟩ ⟨zd⟩ ⟨zt⟩
⟨zh⟩ ⟨zhc⟩ ⟨zhd⟩ ⟨zht⟩

Codas

Some of the onsets and codas above seem to violate the phonotactical rules (e.g., voiced and unvoiced consonants appearing together, as in ⟨cz⟩). This is a result of Night Speech’s romanization scheme, and only affects written language. All the clusters above are prounounced in accordance with the phonotactical rules (e.g., ⟨cz⟩ is pronounced /ks/).

Intersyllabic (a.k.a., medial) consonant clusters adhere to the following rules:

  • No more than three consonants
  • The first two consonants must together form a legal coda cluster
  • The last two consonants must together form a legal onset cluster

During affixation, certain phoneme patterns arise which are subject to additional rules. These rules affect both pronunciation and spelling; therefore, they are simpler to explain using romanized examples:

  • In any pattern of C1C2V(V)(C3)C2, drop the first occurrence of C2
    • ⟨nactavtre⟩ → ⟨nacavtre⟩
    • ⟨rysnanai⟩ → ⟨rysanai⟩
    • ⟨umvavtre⟩ → ⟨umavtre⟩
  • In any pattern of ⟨r⟩VC⟨r⟩, drop the final ⟨r⟩ and double the preceding C
    • ⟨dracre⟩ → ⟨dracce⟩
    • ⟨tryzre⟩ → ⟨tryzze⟩
    • ⟨druchre⟩ → ⟨druchhe⟩
    • ⟨drylre⟩ → ⟨drylle⟩
  • In any pattern of C⟨lr⟩, drop the final ⟨r⟩
    • ⟨adhlre⟩ → ⟨adhle⟩
    • ⟨ithlre⟩ → ⟨ithle⟩
  • In any pattern of C⟨l⟩V(V)⟨r⟩, move the ⟨l⟩ to the end of the pattern and drop the ⟨r⟩
    • ⟨adhlyr⟩ → ⟨adhyl
  • In any pattern of C⟨r⟩⟨r⟩, drop the second ⟨r⟩
    • ⟨nutrre⟩ → ⟨nutre⟩
    • ⟨etrre⟩ → ⟨etre⟩
  • In any pattern of ⟨ruyl⟩, drop the ⟨r⟩
    • ⟨nutruyl⟩ → ⟨nutuyl
    • ⟨etruyl⟩ → ⟨etuyl
  • In any pattern of C⟨luyl⟩, drop the first ⟨l⟩
    • ⟨ithluyl⟩ → ⟨ithuyl
  • If an illegal medial consonant cluster arises, drop the second consonant; repeat until the cluster is legal
    • ⟨nactdrylle⟩ → ⟨nacdrylle⟩
    • ⟨umvnactre⟩ → ⟨umnactre⟩

Note that these changes only apply to patterns that appear across syllable boundaries as a result of morphological processes. Uninflected, non-compound words are never affected.

Syllabification

Syllable breaks are placed according to the following rules:

  • Never divide word-initial consonant clusters: ⟨mvar⟩, ⟨dvus⟩
  • Divide medial and word-final clusters of two consonants:
    • Between nasals and any other consonant: ⟨um-vyr⟩, ⟨uv-n⟩
    • Between stops and fricatives at different places of articulation: ⟨ac-zyr⟩, but ⟨u-dzyr⟩
    • Between fricatives and approximants: ⟨us-re⟩
    • Between any combination that is not a legal onset: ⟨az-zhul⟩
  • Divide after the first in a three-consonant cluster: ⟨nac-tre⟩, ⟨um-vre⟩
  • Divide after vowels followed by single consonants: ⟨zy-vyr⟩
  • Never divide diphthongs: ⟨nuar⟩
  • Never divide long vowels: ⟨tsy-ly-zaan⟩
  • Divide between adjacent, non-diphthong vowels: ⟨zi-yr⟩

Stress

Night Speech stress is regular and can be determined by following these steps:

  1. Stress the final syllable: ⟨dhavyr
  2. If the final syllable is a standalone nasal or ends in a vowel, move the stress to the penultimate syllable: ⟨dhavre⟩
  3. If the word contains a doubled consonant, move the stress to the vowel immediately preceding the last instance of such phonemes: ⟨azzytyr⟩
  4. If the word contains a long vowel or diphthong, move the stress to the last instance of such phonemes: ⟨sylvaa