The story of Algonquian
01
From Mission Word Lists to a Proven Family
Europeans met Algonquian languages before any others on the northern Atlantic coast, and the family resemblance was noticed early. French Jesuits in the 17th century worked in Montagnais (Innu) and Algonquin and treated them as closely akin, in sharp contrast to Huron, an Iroquoian language they also learned. English colonists compiled word lists of Massachusett, Narragansett and Powhatan without much thought of comparison.
The first explicit argument came in 1788, when Jonathan Edwards the younger, who had grown up among the Mahican at Stockbridge, Massachusetts, published Observations on the Language of the Muhhekaneew Indians. He set Mahican beside Delaware, Ojibwe, Shawnee and others, showed that they shared a large common vocabulary, and pointed out that Mohawk, spoken close by, shared none of it. Albert Gallatin's survey of 1836 grouped the languages as 'Algonkin-Lenape', and John Wesley Powell's classification of 1891 fixed the name Algonquian and took in the outlying Blackfoot, Cheyenne and Arapaho.
These were groupings by resemblance. The proof that the resemblances were inherited came from Leonard Bloomfield, who between 1925 and 1946 worked out regular sound correspondences among Fox, Cree, Menominee and Ojibwe and reconstructed their common ancestor in detail. Because he began with these four, his Proto-Algonquian was for a time really 'Proto-Central Algonquian'. In the decades after, Ives Goddard, David Pentland, Paul Proulx and others brought in the eastern and Plains languages, showed that the same reconstruction accounted for them, and revised it where they did not fit. Goddard's work on Arapaho and on the Eastern Algonquian languages was decisive in showing that even the most transformed languages of the family follow regular laws.
02
A Homeland Somewhere to the West
Where Proto-Algonquian was spoken is argued from three kinds of evidence, none conclusive.
The first is linguistic palaeontology. In 1967 Frank Siebert gathered the plant and animal names that can be reconstructed for the parent language, among them words for moose, bear and a number of trees, birds and fish, and looked for the area where all those species lived together. He placed the homeland between Georgian Bay and the Ottawa River in southern Ontario. Critics have pointed out that many of the species ranged widely, so that the method narrows the region less than it seems to.
The second is the shape of the family itself. The most divergent languages, Blackfoot, Arapahoan and Cheyenne, are in the west, and Ives Goddard (1994) showed that many innovations run in a cline from west to east across the family. Divergence is usually greatest near where a family began, so this points to a western origin, with the languages of the Great Lakes and the coast descended from later waves of expansion. The Algic relatives Wiyot and Yurok, in northern California, fit the same picture, and some scholars have looked as far as the interior Northwest for Proto-Algic.
The third is archaeology, which has been tied to Algonquian expansion only tentatively. Several archaeologists have linked the eastward spread to changes in pottery and settlement during the Woodland period, from about 1000 BCE, but material culture does not carry a language label. Genetic studies have so far said little on the question; ancient DNA from the region is scarce, and many communities have been cautious about sampling.
Most specialists would date Proto-Algonquian to somewhere around 2,500 to 3,000 years ago, but this is an informed guess, not a measurement.
03
Animacy, Obviation and the Person Hierarchy
Algonquian languages sound quite different from one another. Cree and Ojibwe have compact sound systems with contrasting long and short vowels, and Cheyenne and Arapaho reshaped the inherited sounds almost beyond recognition. The grammar is where the family is most obviously one.
Every noun is either animate or inanimate. People, animals and spirits are animate, but so are some things that seem arbitrary to an English speaker: in Plains Cree, mitâs 'trousers' is animate, as are some fruits and stones. The class shows in plural endings, in demonstratives and above all in the verb, which takes different stems and endings for animate and inanimate subjects and objects.
Verbs carry most of the grammar. In the Plains Cree niwâpamâw 'I see him or her' and kiwâpamin 'you see me', the prefix belongs to whichever participant ranks higher on a hierarchy, second person above first above third, and a separate marker shows who is acting on whom. Linguists call this the direct–inverse system. Among third persons, the one at the centre of attention is 'proximate' and the rest are 'obviative', so a story about a hunter and a bear can say without ambiguity which of them saw the other. The languages also distinguish an inclusive 'we' (you and I) from an exclusive one (we but not you), as in Cree kîyânaw and niyanân.
Because so much is packed into the verb, a single word often does the work of an English sentence, and word order is fairly free, shaped by what is new or emphasised rather than by grammatical role.
The languages were unwritten before contact. Since then they have been written in Latin alphabets, in the Cree syllabary and its adaptations for Ojibwe and Naskapi, and in the Great Lakes syllabary used by Meskwaki, Sauk and Potawatomi.
04
What Is Still Argued Over
The biggest internal question is subgrouping. Eastern Algonquian is accepted as a genuine branch, defined by shared innovations, but the rest of the family is not neatly divided. 'Central' and 'Plains' Algonquian are geographical labels. Some smaller groupings, such as Cree with Innu and Naskapi, Ojibwe with Potawatomi, and Meskwaki with Sauk and Kickapoo, are secure; how they relate to one another is not. Goddard's west-to-east model treats the family as a series of departures and dialect waves, not as a neat branching tree, and many specialists now describe it that way.
Blackfoot is the hardest case. Its vocabulary and grammar are clearly Algonquian, but its sound correspondences remain only partly worked out, and it is not settled whether it split off first or has simply changed more than the others, perhaps through long contact with non-Algonquian neighbours. Arapahoan and Cheyenne raise similar questions in milder form.
Dating is another open point. There is no agreed method for putting a year on Proto-Algonquian, and the usual figure of around 1000 BCE rests on comparison with better-dated families, not on direct evidence. Linking the family's spread to particular archaeological cultures remains speculative.
Beyond Algic, the record is mostly of proposals that did not hold. Edward Sapir's 'Algonkin-Wakashan' (1929) joined Algic to Kutenai and to the Wakashan and Salishan families of the Northwest Coast; almost no one now accepts it. Mary Haas, who settled the Algic question, also suggested in 1958 that Algonquian was related to the Muskogean and other Gulf languages of the Southeast; the idea drew interest but has not been demonstrated. For now, Proto-Algic is as deep as the comparative method reliably reaches.