2.1.  Phonology.  Coeur d’Alene has forty‑two consonants and five vowels.  The Coeur d’Alene sound system is in keeping with the general Salishan phonological system described extensively by Thompson 1979; it is interesting for its voiced obstruents, which are unusual among the Interior languages, and for its coronal pharyngeals and retracted vowels.

 

Coeur d'Alene Consonants
p t   c   č  kʷ q  qʷ ʔ
p̕   t̕   c̕     č̕    k̕ʷ  q̕ q̕ʷ  
b d   ǰ  gʷ x̣ x̣ʷ  
  s    ɬ   š xʷ x̣ʷ   h
m  n l   r y w ʕ ʕʷ  
m̕   n̕   l̕   r̕ y̕   w̕  ʕ̕  ʕ̕ʷ   

The consonants contrast eleven places of articulation: labial, alveolar, alveopalatal, lateral, labiovelar, uvular, labio‑uvular, coronal pharyngeal, pharyngeal, labiopharyngeal, and laryngeal.  Unlike other Interior Salishan languages, there is no plain (unlabialized) velar series.  There are six manners of articulation for the consonants: plain and glottalized voiceless stops and affricates; voiced stops and affricate; voiceless fricatives; and plain and glottalized resonants.  


2.1.1.1.  Obstruents.  The obstruents are produced as stops, fricatives, or affricates at eight points of articulation.
Labials.  /p/ is a voicelss bilabial stop, which contrasts with /p̕/, a voiceless glottalized bilabial, and /b/ a voiced bilabial stop.  There is no labial fricative.
Alveolars.  The alveolar series includes the voiceless /t/ and voiceless glottalized /t̕/, the voiced /d/ and the voiceless fricative /s/, the voiceless affricate /c/ (IPA [ʦ]) and its glottalized counterpart /c̕/ ([ʦ̕]).  The voiceless fricative /ɬ/ is bilateral.  There is no alveolar voiced affricate.  Coeur d’Alene also lacks the glottalized lateral affricate /ƛ̕/ that occurs throughout the most of rest of the Salishan family; in Coeur d’Alene, this has historically merged with /t̕/.
Alveopalatals.  The Coeur d’Alene unvoiced alveopalatal series includes the plain and glottalized affricates /č č̕/ ([ʧ ʧ̕]) and the fricative /š/ ([ʃ]).  The series is identical to that of Spokane and cognate with Colville /k k̕ x/.  The alveopalatal series also includes /ǰ/, a voiced affricate ([ʤ]).
Labiovelars.  The voiceless velar stops /kʷ k̕ʷ/, the voiced velar stop /gʷ/ and the voiceless fricative /xʷ/ are all produced with the tongue blade at or approaching the front of the velum.  The labialization that accompanies these sounds may be produced either by rounding the lips or by spreading the lips.
Uvulars.  The uvular consonants /q, q̕ x̣/ are voiceless sounds produced with the dorsum of the tongue in contact with the uvula.  The fricative /x̣/ ([χ]) is particularly turbulent, often sounding like throat‑ clearing.      
Labio‑uvulars.  The voiceless labio‑uvular stops /qʷ q̕ʷ/ and fricative /x̣ʷ/ are equivalent to the uvular series with concomitant labialization, again produced either by lip rounding or spread.
Laryngeals.  The two laryngeal sounds in Coeur d'Alene are the fricative /h/ and the glottal stop /ʔ/.


2.1.1.2.  Resonants.  The Coeur d'Alene resonants are produced at six points of articulation.
Nasals.  The plain labial and dental nasals /m, n/ are like those of English.  The glottalized nasals /m̕/ and /n̕/ are produced either as a sequence of nasal resonant and glottal release, or with a creaky voice.      
Laterals.  Coeur d'Alene /l/ is apicoalveolar; its glottalized counterpart /l̕/ is, like the other glottalized sonorants, produced either as a sequence of lateral and glottal stop or as a creaky lateral.      
Glides.  The plain palatal and labiovelar glides /y/ ([j]) and /w/ are like those of English.  The glottalized glides /y̕, w̕/ are produced with palatal or velar constriction interrupted by glottal closure.  All may vocalize; see §2.3.4.      
Coronal pharyngeals.  The coronals /r r̕/ are produced with the tongue tip approaching the alveolar ridge and simultaneous weak pharyngeal constriction.  Along with the uvular obstruents and pharyngeal resonants, the coronal pharyngeals trigger regressive lowering of vowels (see §2.3.1).
Pharyngeals and labiopharyngeals.  The noncoronal pharyngeal resonants /ʕ ʕ̕ ʕʷ ʕ̕ʷ/ are produced with the tongue tip and blade in relatively neutral positions (as for IPA [a]), but with the root of the tongue pulled toward the back wall of the pharynx; the walls of the pharynx are also constricted.  Labialization and glottalization are as described for other resonants; however, labialization is often difficult to detect, particularly if the speaker is one who prefers spread‑lip labialization to rounding.

 

2.1.1.3.  Glottalized segments.  Each voiceless stop or affricate has a phonemic glottalized counterpart.  Glottalization of obstruents is produced by simultaneous release of airflow at both the glottis and the point of articulation.  Gottalization is often very light in word-final or phrase-final position, or with the anterior stops.  The following minimal pairs of roots (indicated by √; see discussion of roots §2.2.1) provide evidence of phonemic glottalization:  √piɬ 'scattered', √p̕iɬ 'persons sit'; √xʷet 'exhausted', √xʷet̕ 'hurry'; √ceqʷ 'pink', √c̕eqʷ 'butcher'; √nič 'drive one', √nič̕ 'be cut'; √kʷus 'curly', √k̕ʷus 'easily split'; √qʷes 'blur', √q̕ʷes 'wrinkled'.
The resonants also occur glottalized and unglottalized phonemically:  √t̕im 'shake hands', √t̕im̕ 'tear cloth from bolt'; √gʷar 'scrape', √gʷar̕ 'be silvery'; √qel 'be fresh', √qel̕ 'swing'.  The resonant may be glottalized by rule, as the diminutive (§2.2.4.1), which glottalizes all resonants within a word:  sníneʔ 'owl', sn̕ín̕eʔ 'little owl'.  Glottalized sonorants are produced either with a creaky voice or as a sequence of glottal stop and sonorant.  The order of the sonorant/glottal stop sequence may vary with environment; for example, with /w̕/ the sequence is generally sonorant‑stop in word‑final position, as in ʔecenšəlčeɬníw̕  'they were surrounded', but stop‑sonorant intervocalically, as in lax̣tíw̕es  'they were friends'.      
The Coeur d'Alene voiced stops /b d gʷ/ and affricate /ǰ/ occur only unglottalized.  [gʷ], however, varies with both [w] and [w̕] in certain environments: x̣esísgʷel 'edible fish', p̕at̕áswel from //p̕at̕=isgʷel// 'trout'; gʷent 'it is low', ʔuw̕ént from //gʷ+√gʷen‑t// (plus some element of C1 glottalization) 'it is very low'.

 

2.1.1.4.  A note on voiced obstruents.  The voiced obstruents are apparently recent introductions into the language.  Many of the words containing /b/ are either borrowed or onomatopoeic, and /b/ occurs only in root-initial position (see discussion of roots §2.2.1) .  The few words conataining /b/ that are not clear borrowings or children's words are all of the shape bVm.  /d gʷ ǰ/ have wider distribution, but only /gʷ/ occurs in morphemes other than roots.

 

2.1.2.  Vowels.  Coeur d'Alene has five vowels in a system distinguishing two levels of height and two degrees of backness:

    

i  
u
e  
(ə)
o
 
a
  

 

     /i/ and /u/ are prototypical high vowels.  /e/ ranges fairly freely among and [e], [ɛ], and [æ] with [ɛ] the most common variant (see discussion of vowel length, §2.3.3).  /o/, which ranges from [ɔ] to [o], and /a/ are vowels that include an element of pharyngeal constriction (see discussion of harmony, §2.3.1).  Schwa is never stressed, and occurs as the reduced form of some unstressed vowels (§2.1.4).  Schwa may also occur as an excrescent element to break up consonant cluster; its use varies with the speaker.

 

2.1.3.  Syllables.  Phonological and morphological processes in Coeur d'Alene are dependent on two types of consonant and vowel (CV) structures: the syllable (discussed here), and the phonological root (discussed in §2.2.1).
Coeur d'Alene syllable structure includes two core shapes:  CV, and CVC:  ʔu kʷu nás 'you are wet' (CV CV CVC); ʔul paq heʔúseʔ  'Easter egg' (CVC CVC CV CV‑CVC).
V and VC syllables appear to occur in a few proclitics or prefixes in word‑initial position.  These vowels transcribed in initial position by Reichard (1938) and Nicodemus (1975) are actually preceded by a glottal stop. For example, Reichard's (1938:593) transcription äts‑gwítc‑stus is actually ʔecgʷíčstus 'he sees her'.
Additional consonants may be added to the onset or coda of a CVC syllable, as long as the sonority of segments decreases with distance from the nucleus: with S representing segments that are of relatively higher sonority, these additional syllables are CSV(C) and CVSC.  Examples include twe 'with' (CSV), uɬxʷíst 'he went again' (CV‑CVSC), and y̕aʕ̕pqín̕ 'lots' (CVSC‑CVC).  A third segment may be added to a coda, following the same restriction: scenč̕mcínčt 'wrist' (C‑ CVC‑CC‑CVSSC).
In this last example, the resonant m surrounded by consonants may serve as a nonvowel syllable peak.  Other examples of syllabic resonants (R) include: čn nxʷétp 'I got out of breath' (CR R‑CVC‑C); léǰncelm 'I got stung' (CVC‑R‑CVC‑R).
Many of the Salishan languages have a propensity to consonant clustering which has obscured the simplicity of the syllable structure.  In Coeur d'Alene, tautosyllabic consonant clusters are limited to those already described, and possibly some sequences of s and a following obstruent:  sút̕mstus 'she stretched it' (CVC‑C‑CCVC); sqʷílkʷəp 'match' (CCVC‑ CVC).  In other cases, s preceding an obstruent seems to maintain a unique timing slot unassociated with the following syllable: sn̕ín̕eʔ 'little owl' (C‑CV‑CVC); scanɬax̣pílgʷes 'worry' (C‑CVC‑CVC‑CVC‑CVC).
All other consonantal material is extrasyllabic.  Pronunciation requires each element in a sequence of unsyllabified stops to be fully released, without the introduction of a syllabic element:  ttm̕íx̣ʷ 'bird; small animal' (C‑C-CVC); tgʷel̕  'because; why' (C‑CVC); hnkʷiʔc  'in the night; at night' (CC‑CVC‑C); ʔešétštstmelp 'you folks tease me' (CV‑CVC‑CC‑CCCVCC); sčšápstq 'chasing' (C‑C‑CVC‑CC‑C).
Schwa excrescence may be used by individual speakers to break up consonant clusters.  Women seem to have a stronger tendency to use excrescent vowels.  The excrescent schwa may assimilate to surrounding sounds and take the same form as unstressed vowels.  The processes of sonorant syllabification and schwa excrescence both allow movement through sequences of consonants that would otherwise appear unpronounceable.

 

2.1.4.  Stress.  Any account of Coeur d'Alene stress will require reference to the stress weight or accent of individual morphemes to account for pairs such as the following, where two roots are subject to the same derivation but result in different stress assignments:
      x̣esísgʷel          //√x̣es=isgʷel//         'edible fish'
     péqsgʷel          //√peq=isgʷel//         'halibut (white fish)'

      tč̕əmíčn̕          //t√č̕əm=íčn̕  //         'ridge'

      tp̕íɬeč̕n̕          //t√p̕íɬ=ič̕n̕ //         'folks sitting on side of hill'

 

     There are several categories of Coeur d'Alene morphemes that are never assigned stress:  
     • prefixes
     • pronominal suffixes
     • vowelless morphemes
     • most transitivizing suffixes

Of the transitivizing suffix sequences, only ‑ši‑t benefactive and ‑tuɬ‑t dative can be assigned stress.   Also, the following vowels are never stressed:

     • /e/ of a (lexical) suffix or of the very common morpheme‑final element ‑eʔ:
          q̕ʷádal̕qs          //√q̕ʷed=el̕qs//                      'blackrobe; priest'
          púteʔntxʷ         //√puteʔ‑n‑t‑∅‑xʷ//            'you honor him'
     • schwa (/ə/), no matter the source
     • /u/ and /i/ that result from glide vocalization

 

     The tendency in Coeur d'Alene is to penultimate or final stress or, more accurately, to stress on the penultimate or final stressable vowel:  
     niʔyl̕xʷúsšnn    //niʔ√yel̕xʷ=us=šin‑n//        'apron'
     yl̕xʷáqsn          //√yel̕xʷ=aqs‑n//                  'bib'

Unstressed root and suffix vowels often (but not always) reduce or delete, leaving the impression that both examples carry final stress.  But the terms of stress placement refer to all stressable vowels; reduction and deletion, as well as schwa excresence and glide vocalization, can occur only after stress is assigned.
In sequences of two or more lexical suffixes, it is often the first that is given stress, usually resulting in (ante)penultimate stress:
     niʔɬəqʷípw̕esšən          //niʔ√ɬaqʷ=ip=iw̕es=šin//     'breechclout'
     st̕əqápw̕asqn                //s√t̕aq=ip=iw̕es=qin//         'beard'

 

     In some cases, the longer sequences of lexical suffixes may give up their stress to the root, or stress is held at the penultimate vowel:  
     unyarp̕y̕ígʷtenč            //u hn√yarp̕=iy̕=igʷt=inč// 
     'Lassos were looped all neglected covering the wall.'

 

     These stress patterns have not succumbed to metrical analysis.  Traditionally, accounts of Salishan stress have relied on inherent stress valencies of roots and affixes (see, for example Thompson and Thompson 1992), as will any future account of Coeur d'Alene stress (see, for example, Czaykowska-Higgins 1993).

    The rules of Coeur d'Alene reduction versus deletion of unstressed vowels are currently impenetrable.  Of the examples discussed at the beginning of this section, the form sxʷít̕ečn̕ 'dentalium' (//s√xʷit̕=ičn̕// ) shows that some suffixes may retain an unstressed vowel following stressed roots.  On the other hand, péqsgʷel 'halibut' loses the first vowel of its unstressed suffix =isgʷel  'fish'.  The words səsaríčn̕  'cricket' ( //s‑√sar=ičn̕// ) and x̣esísgʷel  'fish' have stressed suffixes with roots that retain unstressed vowels. In a word like 'huckleberry', however, the root loses its vowel when it is not assigned stress:   st̕šástq  //s√t̕iš=astq// 'huckleberry'

     Schwa occurs as the reduced forms of most unstressed vowels.  [o a e] also occur as reduced variants of unstressed vowels depending upon the features of the original vowel and its labial, pharyngeal or neutral environment. Unstressed /i/ of prefixes may retain its high front position, but will be lax.

 

 

Grammatical sketch

Overview
Phonology
Morphology
Morphophonology
Morphosyntax
Syntax

Bibliography

 

Contact

Ivy Doak
Timothy Montler