Variable: Does the peripheral suffix distinguish animate obviative plural from animate obviative singular and animate proximate plural?
| Values (view key to language codes) | |||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||
| Key to values | |||||||||||||||||||||||||||||||||||
|
Contents:
- Background on obviative number
- Variation in the realization of obviative plural
- Survey of variation
- Discussion of variation
Background on obviative number
The number of an obviative nominal can be expressed in the inflectional slot known as the peripheral suffix (Goddard 1969: 38, 2007: 209). The peripheral suffix, which occurs word-finally in noun forms, non-personal pronoun forms, participles, many independent-order verb forms, and some conjunct-order verb forms, is a fusional marker of gender (animate or inanimate), number (singular or plural), and obviation (proximate or obviative).
According to Goddard (2003: 59), Proto-Algonquian had the three formally distinct sets of peripheral suffixes shown in Table 1, each set expressing the same contrasts and sharing the same consonantism.
Peripheral suffixes in Proto-Algonquian
| Category | Peripheral suffix sets | ||
|---|---|---|---|
| a-suffixes | i-suffixes | e-suffixes | |
| animate proximate singular | *-a | *-a | *-aː(kaː) |
| animate proximate plural | *-aki | *-iki | *-akeː |
| animate obviative singular | *-ari | *-iri | *-areː |
| animate obviative plural | *-ahi | *-ihi | *-aheː |
| inanimate singular | *-i | *-i | *-eː |
| inanimate plural | *-ari | *-iri | *-areː |
The distribution of the three sets is conditioned by the category of the base:
- Nouns and independent indicative verbs take a-suffixes.
- Participles and certain independent modal forms take i-suffixes.
- Non-personal pronouns, such as demonstratives and indefinite pronouns, may incorporate an a-suffix, an i-suffix, or an e-suffix.
Across most of the family, the animate peripheral suffixes are the only markers in the entire inflectional system that can express the category of obviative plural (*-ahi). This category is not normally expressed by the inanimate peripheral suffixes, which are unmarked for obviation in most languages, nor by the obviative markers that appear in other inflectional slots, which are all number-neutral: the supplementary verbal theme sign *-em ‘obv’, the supplementary verbal central suffix *-ri ‘obv’, and the nominal central suffix *-riw ‘obv possessor’.
Variation in the realization of obviative plural
Although the animate peripheral suffix can express the category of obviative plural, it does not always do so. Across the languages, three patterns in the realization of obviative plural are observed:
- Full contrast: There is an animate obviative plural peripheral suffix that is distinct from both the animate obviative singular peripheral suffix and the animate proximate plural peripheral suffix.
- Number not distinguished for obviatives: The same peripheral suffix is used for animate obviative plural and animate obviative singular. (Result: animate obviatives are number-neutral.)
- Obviation not distinguished for plurals: The same peripheral suffix is used for animate obviative plural and animate proximate plural. (Result: animate plurals are obviation-neutral.)
The three patterns are shown schematically in Table 2, using the shapes of the Proto-Algonquian a-suffixes for illustrative purposes.
Patterns of realization of animate obviative plural
| Pattern | an obv sg | an obv pl | an px pl |
|---|---|---|---|
| Full contrast | -ari | -ahi | -aki |
| Number not distinguished for obviatives | -ari | -aki | |
| Obviation not distinguished for plurals | -ari | -aki | |
Survey of variation
The pan-Algonquian distribution of the three patterns, summarized in the map at the top of this page, is given in greater detail in Table 3, which builds upon an earlier survey in Bliss and Oxford 2017. The underlying shapes of the relevant a-suffixes are provided for each language in the table. The less frequent i-suffixes and e-suffixes normally show the same pattern of syncretism as the a-suffixes.
Survey of obviative plural realization
| Language | an obv sg | an obv pl | an px pl |
|---|---|---|---|
| Proto-Algonquian | *-ari | *-ahi | *-aki |
| Meskwaki-Kickapoo | -ani | -ahi | -aki |
| Sauk | -ani | -aki | |
| Miami-Illinois | -ali | -ahi | -aki |
| Shawnee | -ali | -hi | -aki |
| Old Algonquin | -ar | -aʔ | -ak |
| Ojibwe I | -an | -aʔ | -ak |
| Ojibwe II | -an | -ak | |
| Potawatomi | -ən | -ək | |
| Menominee | -an | -ak | |
| Western Cree | -a | -ak | |
| East Cree | -h | -ič | |
| Naskapi | -a | -ič | |
| Innu | -a | -at | |
| Mi’kmaq I | -l | -Ø | -k |
| Mi’kmaq II | -l | -k | |
| Maliseet-Passamaquoddy | -əl | -ə | -ək |
| Eastern Abenaki | -al | -a | -ak |
| Western Abenaki | -a | -ak | |
| Massachusett, Loup | -ah | -ak | |
| Western Mahican | -ah | -ak | |
| Eastern Mahican | -an | -ah | -ak |
| Delaware (Munsee, Unami) | -al | -ak | |
| Powhatan | -ah | -ak | |
| Cheyenne | -o∼-óhó | -o | |
| Arapaho | -Ø | -o | -oʔ |
| Gros Ventre | -ɔ | -ɔh | |
| Blackfoot | -i | -iksi | |
The sources that were consulted for each language are as follows, listed in the same order in which the languages appear in the table (Central, Eastern, Plains):
- Proto-Algonquian: Bloomfield 1946: 95; Pentland 1999: 244; Goddard 2007: 209.
- Meskwaki: Bloomfield 1927: 182; Goddard 2004: 104.
- Kickapoo: Voorhis 1974: 26.
- Sauk: Whittaker 1996: 385.
- Miami-Illinois: Costa 2003: 271.
- Shawnee: Voegelin 1936: 397; Parks 1975: 136.
- Old Algonquin: Daviault 1994: 379, 386.
- "Ojibwe I" includes many varieties in Saskatchewan, Manitoba, and northwestern Ontario (Valentine 1994: 552), including Weagamow Oji-Cree (Rogers 1964: 119).
- "Ojibwe II" includes Southwestern Ojibwe (Nichols 1980: 185), Nishnaabemwin (Bloomfield 1957: 39; Valentine 2001: 211), Nipissing Algonquin (Jones 1977: 49), and Northern Algonquin (Brouillard & Dumont-Anichinapeo 1987: 66).
- Note: For Algonquin and Ojibwe, the <k> in the table represents a lenis /k/, as in the orthography used in Bloomfield 1957 and Nichols 1980; in other orthographies this is written as <g>.
- Potawatomi: Hockett 1966: 63.
- Menominee: Bloomfield 1962: 113.
- Western Cree: Plains Cree (Wolfart 1973: 43), Woods Cree (Starks 1992: 39–41), Swampy and Moose Cree (Ellis 1971: 87), Atikamekw (Béland 1978: 26).
- East Cree: Junker et al. 2012: 25–26.
- Naskapi: Brittain 2001: 302.
- Innu: Drapeau 2014: 39.
- "Mi’kmaq I": Proulx 1978: 74; Francis & Hewson 2016: 35.
- "Mi’kmaq II": exceptional pattern in certain noun stem classes (Proulx 1978: 75).
- Maliseet-Passamaquoddy: LeSourd 1993: 18.
- Eastern Abenaki: Voorhis 1979: 16.
- Western Abenaki: Prince 1901: 351; Goddard 2008: 288.
- Massachusett: Goddard & Bragdon 1988: 512.
- Loup: Gustafson 2000: 44–46, 81.
- Mahican: Goddard 2008: 288, 293.
- Delaware (Munsee and Unami): Goddard 1969: 154.
- Powhatan: Siebert 1975: 418.
- Cheyenne: Goddard 2000: 80–83.
- Arapaho: Cowell & Moss 2008: 54–57.
- Gros Ventre: Cowell 2013: 38.
- Note: Arapaho and Gros Ventre have complex morphophonemics; the forms given in the table are those that appear with the most common class of noun stems.
- Blackfoot: the forms given in the table are those that appear on nouns (Goddard 2018: 84); verbs show -ini ‘an obv sg’ and -i ‘an pl’ (Goddard 2018: 90–91).
Discussion of variation
Jump to: Prevalence of patterns | Parallel independent developments | Language-internal variation | Direction of levelling
Prevalence of patterns
The family shows a roughly even split between the conservative full contrast pattern and the number syncretism pattern (number not distinguished for obviatives), with instances of both patterns attested within each of the traditional Central, Eastern, and Plains groupings.
The obviation syncretism pattern (obviation not distinguished for plurals) is much less frequent, occurring in only three languages that are each rather innovative in other ways as well (Mi'kmaq, Gros Ventre, Blackfoot).
Parallel independent developments
The number and obviation syncretisms both developed more than once. The number syncretism developed at least once in the west (in Cree, Ojibwe-Potawatomi, Menominee, and Cheyenne) and at least once in the east (in Western Abenaki, Southern New England Algonquian, Western Mahican, Delaware, and Powhatan). In both locales, it is likely that there were in fact multiple independent developments, as indicated by the differences in direction of levelling discussed below. The obviation syncretism likewise developed at least once in the west (in Blackfoot and Gros Ventre) and once in the east (in Mi’kmaq). Since these syncretisms are so widespread and evidently emerge so readily, they do not diagnose fundamental divisions in the family.
Language-internal variation
The shallowness of the variation in the realization of obviative plural is underscored by the existence of such variation even within genetic groupings that are themselves quite shallow. In Ojibwe-Algonquin, southern varieties show the number syncretism but many northern varieties retain the full contrast. In Meskwaki-Sauk-Kickapoo, Sauk shows the number syncretism while Meskwaki and Kickapoo retain the full contrast. And in Delaware-Mahican, Eastern Mahican retains the full contrast but Western Mahican and Delaware both developed a number syncretism and likely did so independently of each other (Goddard 2008: 288–290).
Direction of levelling
From a diachronic perspective, the emergence of the number and obviation syncretisms usually involves levelling in paradigms that use peripheral suffixes: one peripheral suffix is generalized to contexts that originally called for a distinct peripheral suffix. Such instances of levelling can, in principle, differ in their direction: which of the two original markers is generalized at the expense of the other?
In the three languages with an obviation syncretism, in which animate proximate plural and animate obviative plural are marked the same, the direction of levelling is uniform: it is always the original proximate plural suffix (*-aki) that is used as a general animate plural marker.
In the numerous languages with a number syncretism, in which animate obviative singular and animate obviative plural are marked the same, the direction of levelling varies: the general animate obviative marker reflects obviative singular *-ari in some languages and obviative plural *-ahi in others, as discussed by Siebert (1975: 419) and Goddard (1991: 62, 2008: 288–290). Table 4 summarizes the direction of levelling in these languages. Sauk, many varieties of Ojibwe-Algonquin, Potawatomi, Menominee, and the Delaware languages generalize the singular; Western Abenaki, Massachusett, Loup, Western Mahican, Powhatan, and Cheyenne generalize the plural.
Direction of levelling in number syncretism
| Language | obv sg | obv pl |
|---|---|---|
| Proto-Algonquian | *-ari | *-ahi |
| Sauk | -ani → | |
| Ojibwe-Algonquin | -an → | |
| Potawatomi | -ən → | |
| Menominee | -an → | |
| Cree | -a | -a |
| Western Abenaki | ← -a | |
| Massachusett, Loup | ← -ah | |
| Western Mahican | ← -ah | |
| Delaware | -al → | |
| Powhatan | ← -ah | |
| Cheyenne | ← -o |
There is one language in which the lack of an obviative number contrast cannot be definitively attributed to morphological levelling, as a phonological explanation is also possible. In Cree, regular sound change in word-final position would have shortened both obviative singular *‑ari and obviative plural *‑ahi to ‑a (Bloomfield 1946: 93; Goddard 1991: 62). The number-neutrality of the Cree animate obviative marker ‑a thus has three possible origins:
- Singular *‑ari generalized, as in Southern Ojibwe (morphological change: levelling)
- Plural *‑ahi generalized, as in Cheyenne (morphological change: levelling)
- Singular *‑ari and plural *‑ahi both become ‑a due to the loss of final vowels and the subsequent loss of final *r and *h (phonological change)