Dialect Features
/e/
Subject to conditions that are not fully understood, some dialects tend to have /e/ where other dialects have /i/. Examples, contrasting Saik'uz dialect with Nak'albun, are: /dzen/ vs. /dzin/ "day", /-nen/ vs. /-nin/ "face", and /tesch'ih/ vs. /tisch'ih/ "I will shoot".

Final /k/
Subject to conditions that are not fully understood, the southern dialects have a final /k/ in many forms that is lacking in others. Examples, contrasting Saik'uz dialect with Nak'albun, are: /khunek/ vs. /khuni/ "word" and /-t'ak/ vs. /-t'a/ "back".

Palatal Realization of /aih/
In some dialects /aih/ is realized as [aç].

Final Glottal Stop
In the southern dialects, word-final glottal stops present in the Stuart-Trembleur Lake dialect are often, though not always, absent. The precise conditions on their retention remain to be determined.

n->0/_'
In some dialects /n/ is deleted preceding /'/. For example, in Saik'uz dialect, the /n/ of the second person singular subject prefix /in/ is realized in 'inyi "you (1) are eating", but does not appear in duni'alh "you (1) are chewing". Similarly, the perfective prefix /in/ appears in sghanintan "he gave me (long rigid object)", but loses its /n/ in sghani'ai "he gave me (single generic object". In other dialects there is no such deletion rule. For example, in the Nak'albun dialect "you (1) are chewing" is dunin'alh and "he gave me (single generic object)" is sghanin'ai.

Reflex of /uh/
In the relatively recent past (roughly 100 years ago), Carrier had the rhyme /uh/ (in many cases derived from /ukh/). This has changed into /oh/ in some dialects and into /ah/ in others. For example, *keyukh "territory" has become keyoh in some dialects and keyah in others.

/ukw/ vs. /ook/
While some dialects preserve earlier /ook/, in others this has changed, within the past 100 years, to /ukw/. For example, Saik'uz has lhukw for "fish", while Nazkoh has lhook.

Presence of /gl/
There are no inherited /gl/ clusters in Carrier, but in some dialects /gl/ clusters been introduced by borrowing from French. For example, French le clou "nail" has been borrowed as lugloo. In other dialects, /gl/ is replaced by /dl/, which is present in inherited words, so that "nail" is borrowed as ludloo. The correspondance /gl/ to /dl/ therefore does not reflect a sound change. Rather, it reflects the fact that some dialects dealt more conservatively with French loanwords than did others. The dialects with /dl/ appear to be those in which the intensity of contact with French was lowest. This feature is therefore irrelevant to subgrouping.

/i/ After Velar C
In some dialects /i/ has the allophone [ui] after the velar consonants k, k', g, kh and gh. For example, the Paper Birch tree is called [k'i] in the Nak'albun dialect, but [k'ui] in Saik'uz.

Vocalized /l/
When the valence prefix /l/ stands alone, without any preceding prefix, in some dialects it becomes /lh/, while in others it becomes /lhu/. For example, "it (generic) is white" is lhyul in some dialects, lhuyul in others. (That this is really an /l/ valence prefix, not /lh/, is shown by forms such as nulyul "it (n-class) is white".)

/n'/ Present
Some dialects have a glottalized /n'/ in coda position, always in morphologically derived contexts, namely optative and perfective stems of verbs, and possessed stems of nouns. Most dialects lack glottalized nasals entirely. For example, "my song" is shyunin most dialects but shyun' in some. Similarly, "I will see you again" is nanyoost'enin most dialects but nanyoost'en' in some.

/ng/ Present
In some dialects a velar nasal phoneme /ng/ is marginally present, e.g. in Nak'albun dialect 'utsung "meat". In most dialects this phoneme is absent and "meat" is 'utsun. This feature is not a good diagnostic of historical development as /ng/ is marginal in all dialects and in those dialects that have it, shows up in different words. For example, Lheidli dialect has 'utsun for "meat" but has /ng/ in daingnun "summer".

/tl/->/lh/ after /lh/
In some dialects, /tl/ becomes /lh/ following /lh/. For example, in some dialects "he smears him (e.g. with liniment)" is yulhtloh, but in others it is yulhlhoh.

D-Effect Bleeds /tl/->/lh/
In those dialects in which /tl/ becomes /lh/ following /lh/, this rule may or may not be bled by the d-effect. Thus, in the Saik'uz dialect, we have nelhlhus "you (1) kneaded", but nedultlus "we (2) kneaded", where the final /d/ of the first person dual subject marker changes the /lh/ valence prefix to /l/ and in so doing bleeds the spirantization of /tl/. In contrast, in the Tsetl'adak dialect, we have yulhlhah "he is smearing it" and idullhah "we (2) are smearing". Here the 1d subject marker converts the underlying lh valence prefix to l. but not before the stem tlah becomes lhah.

Long /a/
In some dialects, there are words distinguished only by the length of the vowel /a/. For example, in Cheslatta dialect, khunawhelnuk means "he confessed", while khuna:whelnuk means "they confessed". Vowel length need not be marked in underlying representation. All long vowels are morphophonemically derived. In the above example, the long a: results from the contraction of an underlyingly short a and the /h/ that marks the third person duo-plural subject. This is the source of most long vowels. In most dialects, there is no such distinction of short versus long /a/. (Most if not all dialects have instances of long /i/ due to the fusion of two adjacent short /i/s.) Vowel length in Nadleh, at least in forms like nucha:teskel, is obsolete. Current speakers use uncontracted forms, but remember elders now gone as using contracted forms with long vowels.

1s Possessive Epenthesis
In all dialects the first person singular possessive prefix is arguably underlyingly s- except before class 2 nouns (in general, those beginning with glottal stop), where it is se-. In some dialects, in most circumstances, this s is realized as such, without an intervening vowel, before consonants. Thus, in the Nak'albun dialect we have stoo "my water", sbat "my mittens", {\itskechun "my legs", etc. In other dialects, an epenthetic /u/ is always inserted between the prefix and the stem. This, in the Lhk'acho dialect we have sutoo "my water", suchachebat "my mittens", sukechun "my legs", etc.

1s subj /s/ + valence /lh/
When the first person singular subject prefix /s/ combines with the valence prefix /lh/, the result in some dialects is /s/, in others /lh/. For example, in all dialects "he is waiting for you" is mbalh'i. In some dialects, "I am waiting for you" is mbas'i; in others it is mbalh'i.

1s subj /s/ + valence /l/
When the first person singular subject prefix /s/ combines with the valence prefix /l/, the result in some dialects is /z/, in others /lhu/. For example, in all dialects "she is pregnant" is ulchan. In some dialects, "I am pregnant" is uzchan; in others it is ulhuchan.

Negative /s/ + Valence /l/
When the conjunct negative prefix /s/ combines with the valence prefix /l/, the result in some dialects is /lh/, in others /lhu/. For example, in Saik'uz dialect "he is not running around" is /nulhugaih/, while in the Nak'albun dialect this is /nulhulhgaih/. These forms differ also in that the Saik'uz form has no disjunct negative prefix, only the conjunct negative prefix /s/ that fuses with valence /l/, while the Nak'albun form has both the conjunct negative prefix /s/ and the disjunct negative prefix /lh/.

Perfective /s/ + Valence /l/
When the perfective prefix /s/ combines with the valence prefix /l/, the result in some dialects is /lh/, in others /lhu/. For example, in Saik'uz dialect "we (3+) ran around" is /nuts'ulhghaz/, while in the Tsetl'adak dialect this is /nuts'ulhughaz/.

2dp /h/ + Valence /l/
When the second person duo-plural subject prefix /h/ combines with the valence prefix /l/, the result in some dialects is /lh/, in others /lhu/. For example, in some dialects "you (2+) are pregnant" is ulhchan; in others, ulhuchan.

uh$ -> /o/ or /a/
The rhyme uh never appears as such except when immediately preceding the verb stem. In some dialects it becomes o, while in others it becomes a. For example, in the Nak'albun dialect, "they are looking for me" is: skonuta, while in the Saik'uz dialect, this is realized as skanuta. In both cases the surface form is derived from /skuhnuta/, where the /h/ is the third person duo-plural subject marker.

Optative oo/w Alternation
In some dialects the optative prefix is always realized as a vowel. It is realized either as /oo/ or, when it fuses with a following /i/ and in certain other circumstances, as /o/. In other dialects there is an alternation between these vocalic forms and /wu/. /oo/ or /o/ occurs when a consonant precedes, while /wu/ occurs when there is no preceding consonant. For example, in the Nak'albun dialect, the first person singular optative of "to drink" is oosnai', which becomes 'oosnai' when the unspecified object prefix '- is added. In contrast, in the Saik'uz dialect, the specified object form is wusnai, while the unspecified object form is 'oosnai.

D-Effect on /gh/
The effect that the underlying final /d/s of the first person dual subject prefix and the d-valence prefix have on the following consonant is known as the D-Effect. In some dialects, there is a D-Effect on /gh/. For example, in the Nak'albun dialect, "he is hairy" is dughai, but in digai "we (2) are hairy" the stem-initial consonant changes into /g/ under the influence of the first person dual subject prefix /id/. In other dialects there is no D-Effect on /gh/. For example, in the Saik'uz dialect "he is hairy" is dughai, but "we (2) are hairy" is didughai, not *didugai.

Epenthesis To Realize Valence /d/
In some dialects the valence prefix /d/ is deleted when the following verb stem begins with a consonant with which it cannot fuse via the D-Effect. Thus, in the Nak'albun dialect we have forms like bedugut "saw", that is "the thing by means of which it is sawed", where the d-valence prefix is not directly realized but converts the underlying /gh/ to /g/. In dialects in which there is no D-effect on /gh/, an an epenthetic vowel u is inserted between the valence d and the verb stem. Thus, in the Saik'uz dialect, the same form is bedudughut.

/wh/ + /k/
In all dialects when /wh/ is followed by a velar consonant (k, g, k', kh or gh), the wh becomes h and the velar consonant becomes rounded (kw, gw, kw', khw or ghw). In some dialects the epenthetic u becomes oo in this context, while in others it does not. For example, in the Nak'albun dialect "I am looking for (a house)" is hukwunusta, derived from whu-kunusta, while in the Saik'uz dialect it is hookwunusta.

2s /in/ -> n/#_
In some dialects the /i/ of the second person singular subject prefix /in/ is not realized if it is not preceded by another prefix. Thus, in the Nazkoh dialect "you (1) are eating something" is 'inyi, where the unspecified object prefix ' causes the i to be realized, but "you (1) are eating" is nyi, with the i deleted in the absence of a preceding prefix. In other dialects the realization of the i does not depend on the presence of a preceding prefix.

3s PPobj Labial Shift
In some dialects the third person singular non-obviative object of a postposition before a consonant is oo regardless of the point of articulation of the consonant. In other dialects, if the consonant is non-coronal, no prefix is overtly realized, but if the consonant is velar, it becomes labialized. For example, k becomes kw, and gh becomes w. This may be regarded as absorption of the rounding of the oo by the following consonant, with null realization in the case of labials and overt realization in the case of velars.

PA 1s subject
In all dialects in most circumstances the first person singular subject prefix is /s/. In some dialects this is the case in the perfective affirmative as well. In others, there is a special allomorph /i/ that is used in the perfective affirmative. Thus, in some dialects "I shot" is susch'i, while in others it is sich'i. The suppletive form is clearly an archaism so having it is a shared retention. The extension of the more general form is a simple and obvious regularization, which could easily have occurred independently in different dialects. Indeed, there is direct evidence that it did, since the change is in progress in the Lheidli dialect. Neither value of this feature is therefore of use for subgrouping.

1d subject
In some dialects the first person dual subject prefix is underlyingly /id/, e.g. nidliz "we (2) stew". In other dialects it is /idud/, e.g. nidudlez. The longer form appears to be an innovation, so this is probably a shared innovation characterising the Southern dialects.

3s Possessive Before C
In some dialects the third person singular possessive prefix, and similarly, the third person singular object marker, is realized as oo before consonants, e.g. oogan "his arm". In other dialects it is realized as bu, e.g. bugan.

3dp Possessive Before C
In some dialects the third person duo-plural possessive prefix, and similarly, the third person duo-plural object marker, is realized as bu before consonants, e.g. bugan "their arms". In other dialects it is realized as hubu, e.g. hubugan.

/a/ Future Exists
In some dialects verbs form the future tense in two ways. One class of verb has the tense/aspect vowel i or e.\enumfootnote{This fuses with the initial i or e of other morphemes to yield a in certain forms. The other class of verb has the tense/aspect vowel a.\enumfootnote{This fuses with the initial i or e of other morphemes to yield a in certain forms. In such a dialect, we therefore have forms like tesch'ih "I will shoot", with e, but ootaskulh "I will buy", with a. In other dialects there is only one type of future, marked by e; the a future does not exist. In such dialects "I will buy" is ooteskulh.

Disjunct Negative Prefix
Negation may be marked by either a disjunct prefix, a conjunct prefix, or both. For example, in the Nak'albun dialect, "I am not singing" is lhuzusjun, where lh is the disjunct negative prefix and z is the conjunct negative prefix. Some dialects do not have any disjunct negative prefix. Other dialects have one, but it is different from the Nak'albun prefix.

Modal /gh/
In most dialects, the gh mode prefix of proto-Athabaskan has moved into qualifier position. Thus, for example, we have Saik'uz {\itbutghaghutesnalh "I am going to cook", where the second gh is the historical mode prefix but now precedes the future/inceptive t. In some dialects, however, this prefix has remained in the mode position. Thus, in Cheslatta dialect the cognate form for "I am going to cook" is butghatughesna.

/in/ precedes Si
In all dialects there is a class of verbs that takes the prefix in in the imperfective. In all dialects the /i/ part of this prefix is only realized if it is preceded by a conjunct prefix. In some dialects this prefix competes with the inner subject prefix \Si\ and is only present when there is no \Si\ prefix, that is, in the third person forms and the first person plural. For example, in the Nak'albun dialect the prefix is present in ts'inyiz "we (3+) are tall", but not in usyiz "I am tall". In other dialects, this prefix does not compete with \Si\ but precedes it, so that it is present in all forms. For example, in the Nazkoh dialect "we (3+) are tall" is ts'inyez and "I am tall" is nusyez.\enumfootnote{The prefix is realized as /n/ rather than /in/ since it is not preceded by a conjunct prefix.

1d Possessor/Object = 1p
In some dialects first person dual possessors are not distinguished from first person plural possessors. In the Nak'albun dialect, for example, nedustl'us "our book" refers to a book belong either to two of us or to three or more of us. In other dialects, there is a distinction. For example, in the Chestl'ada dialect "our (2) book" is nahdustl'us while "our (3+) book" is nedustl'us. The same is true of the object prefixes on verbs.

1d/1p Subject Overlap
In most dialects, first person dual and first person plural subjects are distinct. In all dialects the old indefinite subject prefix ts' has extended its range to include the first personal plural. In most dialects, the old first person duo-plural prefix id(ud) has become restricted to the dual, supplanted in the plural by ts'. However, in at least one dialect, this process is incomplete. While ts' can be used only in the plural (or as an indefinite), idud is ambiguous between first person dual and plural.

Instrumental Nouns Conjugated
In all dialects instrumental nouns may be derived from verbs by adding the prefix be-, sometimes accompanied by the suffix -i. In some dialects many such derived nouns are internally conjugated. Thus, in the Nak'albun dialect, "vehicle" is benugoo-i "that which one drives", "my vehicle" is benusgoo-i "that which I drive", "our vehicle" is benuts'ugoo-i "that which we drive", and so forth. Such deverbal nouns may not be marked for possession like ordinary nouns. In other dialects, instrumental deverbal nouns are never internally conjugated and are marked for possession like other nouns. For example, "my vehicle" is sbenugoo-i.

Productive Nominal Plural
In some dialects there is a productive nominal plural suffix -ne added to nouns denoting human beings. Thus, in the Nak'albun dialect there are plurals for relatively recent loanwords, such as gadulek-ne "Catholics" and even for unassimilated foreign words, such as government-ne "government people". In other dialects, there are either no distinct plural forms at all, or just some non-productive, irregular ones.

Plural Suffixes < /yaz/ etc.
In some dialects the plural suffix ne follows suffixes such as yaz "little", cho "big", and ti "big". Thus, we have duneyaz "boy", duneneyaz "boys". In other dialects the plural suffix follows these suffixes, so that "boys" is duneyazne.

Comparative Verb Form
In some dialects simple descriptive verbs like "be big", "be good" have a special comparative form even for simple comparisons. Thus, in the Nak'albun dialect, we have sanus 'ilyiz "he is taller than me". In other dialects, simple comparatives do not take on such a special form. In Cheslatta, for example, the same sentence is sanus nyez.

Agentive Nominalizing Suffixes
In some dialects, nouns meaning "the person who Vs", "the people who V", and "the thing that Vs" may be derived from verbs by adding an overt suffix, e.g. Nak'albun nudaih-un "dancer", from nudaih "he dances". In other dialects, there are no such nominalizing suffixes. Such nominalizations exist, but only as 0-nominalizations.

Locative Nominalizing Suffixes
In some dialects, nouns meaning "the place where something is done" may be derived from verbs by adding an overt suffix, e.g. Nak'albun hunudaih-un "dance hall", from hunudaih "they dance". In other dialects, there are no such nominalizing suffixes. Such nominalizations exist, but only as 0-nominalizations.

Admirative Valence
All dialects permit the derivation of "admirative" forms from many descriptive verbs. For example, in the Nak'albun dialect, we have ntsool "it is small" with the corresponding admirative lhe'ultsool "how small it is!". In all dialects the formation of the admirative involves the addition of the prefix lhe-. In addition, in some dialects, the valence prefix becomes /l/, while in others it becomes /lh/. Thus, in the Saik'uz dialect "how small it is" is lhelhtsool.

Distinct Subject RCs
In some dialects, relative clauses follow the head noun, and the verb of the relative clause optionally takes a suffix that agrees with the head. For example, in the Nak'albun dialect, "I shot the bear that was eating berries" is: Sus mai nuldilh-i susch'i. Here, the suffix -i is the non-human non-areal relativizing suffix and agrees with sus "bear". In other dialects, subject relatives like this one are made differently, and are not distinct from "while" clauses. For example, in the Saik'uz dialect, this sentence is: Sus mai nuldelh hoh susch'i. This can also be interpreted as meaning: "I shot the bear while it was eating berries."

Relativizing Suffixes
In some dialects, the verb of a relative clause may take a suffix that agrees with the head. For example, in the Nak'albun dialect, in the relative clause Mai' oonanyin-i nincha. "The berries that you picked are big", the verb oonanyin bears the suffix -i, which agrees with the non-human non-areal head mai "berries". In other dialects, no such suffix exists.

NP Complement Suffix
In some dialects certain types of complement clause optionally take a nominalizing suffix on the verb. For example, in the Nak'albun dialect,"I like to sing" is usjun-un hoonust'i', in which -un is suffixed to the verb "I sing". In other dialects, no such nominalizing suffix may be used. For example, in the Saik'uz dialect, the only grammatical version of this sentence is: usjun hoonust'i'.

Aspect in Verb Classification
There are certain constructions in which the choice of auxiliary verb depends on the nature of the main verb. For example, in the Nak'albun dialect, "He keeps on singing" is: Ujun za 'utni, with auxiliary verb 'utni, while "He keeps on dancing" is Nudaih za 'ut'en, with auxiliary verb 'ut'en. In some dialects, the aspect of the lower verb plays a role in determining which auxiliary verb is used. Thus, in the Nak'albun dialect, "He keeps on kneading dough" is Lhes nulhtlus za 'ut'en, but "He keeps on pushing it" is Yunilht'i za yulh'en. The choice between 'ut'en and yulh'en is determined by the fact that the first example has punctual aspect while the second example has continuous aspect. In other dialects, aspect does not play a role in determining the choice of auxiliary verb. For example, in the Saik'uz dialect, "He keeps on kneading dough" is Lhes yunulhlhus ze yulh'en, with the auxiliary yulh'en.

Mode With "want"
In some dialects the verb of the complement of "want" may be either in the future or the optative. Thus, in the Nak'albun dialect, "I want to eat" may be either: 'Utis'ulh hukwa'nuszun with the lower verb in the future, or 'Oos'alh hukwa'nuszun with the lower verb in the optative. In other dialects, only the future is permitted. For example, in the Saik'uz dialect, the only grammatical version of this sentence is: 'Utesyilh hookwa'nuszun with the lower verb in the future. *'Oosyi hookwa'nuszun, with the lower verb in the optative, is ungrammatical.

Comparison of Minority
In Carrier the pivot of comparison, equivalent to English "than", is a postposition which, when its object is a pronoun, is inflected as if for possession. Thus, in the Nak'albun dialect, "He is taller than me." is Sanus 'ulyiz., where the first word means "(more) than me" and the second is the comparative form of "he is tall". In some dialects this is the only pivot of comparison. In others, there is a second pivot of comparison with the meaning of "less than". Thus, in the Nak'albun dialect, "He is less tall than me." is Sk'elh'ih 'ulyiz..

Topic Particle
In most Carrier dialects, the topic of the sentence is marked by the use of a resumptive pronoun. That is, the topicalized noun phrase is followed by an appropriate independent pronoun. For example, 'Uba 'en cha soo 'uzus 'ulh'en. "As for Father, he too makes good hides." This contrasts with the untopicalized version: 'Uba cha soo 'uzus 'ulh'en. "Father too makes good hides." When the topicalized constituent is first or second person and so is marked on the verb, the mere use of an independent pronoun has the effect of topicalization. For example, in the Stuart Lake dialect, we could say Si, soo 'ust'oh. "As for me, I am well." In some dialects, however, topicalization is marked by a following topic particle, which takes the form 'en, homophonous with the third person singular human pronoun, which must be its historical source. An example in Nazko dialect is: Si 'en soo 'unust'oh. "As for me, I am fine." Here si is the pronoun "I", soo is "well", and 'unust'oh is "I am". This type of topicalization is found in Slave, as well as in Japanese and Korean.

HOME BACK

Yinka Déné Language Institute © 2006