Southern Athabaskan languages

Family division

The seven Southern Athabaskan languages can be divided into two groups according to the classification of Harry Hoijer: (I) Plains and (II) Southwestern. Plains Apache is the only member of the Plains Apache group. The Southwestern group can be further divided into two subgroups (A) Western and (B) Eastern. The Western subgroup consists of Western Apache, Navajo, Mescalero, and Chiricahua. The Eastern subgroup consists of Jicarilla and Lipan.

I. Plains (AKA Kiowa–Apache)

II. Southwestern

A. Western B. Eastern

Hoijer's classification is based primarily on the differences of the pronunciation of the initial consonant of noun and verb stems. His earlier 1938 classification had only two branches with Plains Apache grouped together with the other Eastern languages (i.e. with Jicarilla and Lipan).

Mescalero and Chiricahua are considered different languages even though they are mutually intelligible (Ethnologue considers them the same language). Western Apache (especially the Dilzhe'e variety) and Navajo are closer to each other than either is to Mescalero/Chiricahua. Lipan Apache and Plains Apache are nearly extinct (in fact Lipan may already be extinct). Chiricahua is severely endangered. Mescalero, Jicarilla, and Western Apache are considered endangered as well, but fortunately children are still learning the languages although the number of child speakers continues to decline. Navajo is one of the most vigorous North American languages, but use among first-graders has declined from 90% to 30% in recent years (1998 N.Y. Times, April 9, p. A20).

Sounds

All Southern Athabaskan languages have somewhat similar phonologies. The description below will concentrate mostly on Western Apache. One can expect minor variations of this description in other related languages (such as Navajo, Jicarilla, Chiricahua).

Consonants

Southern Athabaskan languages generally have a consonant inventory similar to the set of 33 consonants below (based mostly on Western Apache):

Only Navajo and Western Apache have glottalized nasals.

Orthography (consonants)

The practical orthography corresponds to the pronunciation of the Southern Athabaskan languages fairly well (as opposed to the writing systems of English or Vietnamese). Below is a table pairing up the phonetic notation with the orthographic symbol:

Some spelling conventions:

1. Fricatives [h] and [x] are both written as h. (see also #2 below)
2. The fricative [x] is usually written as h, but after o it may be written as hw, especially in Western Apache (may be pronounced [xʷ]).
3. The fricative [ɣ] is written gh the majority of the time, but before i and e it is written as y (& may be pronounced [ʝ]), and before o it is written as w (& may be pronounced [ɣʷ]).
4. All words that begin with a vowel are pronounced with a glottal stop ’ [ʔ]. This glottal stop is never written at the beginning of a word.
5. Some words are pronounced either as d or n or nd, depending on the dialect of the speaker. This is represented in the consonant table above as [ⁿd]. The same is true with b and m in a few words.
6. In many words n can occur in a syllable by itself in which case it is a syllabic [n̩]. This is not indicated in the spelling.

Vowels

Southern Athabaskan languages have four vowels of contrasting tongue dimensions (as written in a general "practical" orthography):

These vowels may also be short or long and oral (non-nasal) or nasal. Nasal vowels are indicated by an ogonek (or nasal hook) diacritic ˛ in Western Apache, Navajo, Mescalero, and Chiricahua while in Jicarilla the nasal vowels are indicated by underlining the vowel. This results in sixteen different vowels:

IPA equivalents for Western Apache oral vowels:

i = [ɪ], ii = [iː], e = [ɛ], ee = [ɛː], o = [o], oo = [ʊː], a = [ɐ], aa = [ɑː].

In Western Apache, there is a practice where orthographic vowels o and oo are written as u in certain contexts. These contexts do not include nasalized vowels, so nasal u never occurs in the orthography. This practice continues into the present (perhaps somewhat inconsistently).

However, in Harry Hoijer and other American linguists' work all o-vowels are written as o. Similarly, Navajo does not use orthographic u, consistently writing this vowel as o.

In Chiricahua and Mescalero, this vowel is written as u in all contexts (including nasalized ų).

Other practices may be used in other Apachean languages.

Tone

Southern Athabaskan languages are tonal languages. Hoijer and other linguists analyze Southern Athabaskan languages as having 4 tones (using Americanist transcription system):

high (marked with acute accent ´, Example: á)

low (marked with grave accent `, Example: à)

rising (marked with háček ˇ, Example: ǎ)

falling (marked with circumflex ˆ, Example: â)

Rising and falling tones are less common in the language (often occurring over morpheme boundaries) and often occur on long vowels. Vowels can carry tone as well as syllabic n (Example: ń).

The practical orthography has tried to simplify the Americanist transcription system by representing only high tone with an acute accent while leaving low tone unmarked:

high tone: á

low tone: a

So now niziz is written instead of the previous nìzìz.

Additionally, rising tone on long vowels is indicated by an unmarked first vowel and an acute accent on the second, vice versa for falling tone:

rising: aá (instead of Americanist: ǎ·)

falling: áa (instead of Americanist: â·)

Nasal vowels carry tone as well, resulting in a two diacritics on vowels with high tone: ą́ (presenting problems for computerization). Recently, de Reuse (2006) has found that Western Apache also has a mid tone, which he indicates with a macron diacritic ¯, as in ō, ǭ. In Chiricahua, a falling tone can occur on a syllabic n: n̂.

Here are some vowel contrasts involving nasalization, tone, and length from Chiricahua Apache:

cha̧a̧  'feces' chaa  'beaver' shiban  'my buckskin' shibán  'my bread' bik’ai’  'his hip' bík’ai’  'his stepmother' hah’aał  'you two are going to chew it' hah’ał  'you two are chewing it'

Comparative phonology

The Southern Athabascan branch was defined by Harry Hoijer primarily according to its merger of stem-initial consonants of the Proto-Athabascan series *k̯ and *c into *c (in addition to the widespread merger of *č and *čʷ into *č also found in many Northern Athabascan languages).

Hoijer (1938) divided the Apachean sub-family into an Eastern branch consisting of Jicarilla, Lipan, and Plains Apache and a Western branch consisting of Navajo, Western Apache (San Carlos), Chiricahua, and Mescalero based on the merger of Proto-Apachean *t and *k to k in the Eastern branch. Thus, as can be seen in the example below, when the Western languages have noun or verb stems that start with t, the related forms in the Eastern languages will start with a k:

He later revised his proposal in 1971 when he found that Plains Apache did not participate in the *k̯/*c merger to consider Plains Apache as a language equi-distant from the other languages, now called Southwestern Apachean. Thus, some stems that originally started with *k̯ in Proto-Athabascan start with ch in Plains Apache while the other languages start with ts.

Morris Opler (1975) has suggested that Hoijer's original formulation that Jicarilla and Lipan in an Eastern branch was more in agreement with the cultural similarities between these two and the differences from the other Western Apachean groups. Other linguists, particularly Michael Krauss (1973), have noted that a classification based only on the initial consonants of noun and verb stems is arbitrary and when other sound correspondences are considered the relationships between the languages appear to be more complex. Additionally, it has been pointed out by Martin Huld (1983) that since Plains Apache does not merge Proto-Athabascan *k̯/*c, Plains Apache cannot be considered an Apachean language as defined by Hoijer.

Other differences and similarities among the Southern Athabaskan languages can be observed in the following modified and abbreviated Swadesh list: