Camas Etymology

The Inland Etymology of Camas

Alan H. Hartley, October 2001

The etymology usually given for the English word camas (also quamash), rooted firmly in reference books since the mid-19th century, is due for an overhaul. It is wrong in its fundamental assertions, namely, that the English comes from Chinook Jargon (CJ) kamas and that the CJ form derives from a Nootka word referring to some sort of fruit (with its implied sweetness): the maritime Nootkan etymology should yield to the inland Sahaptian.

John Jewitt, armorer of the merchant ship Boston, lived as a slave among the Nootkas of Vancouver Island from 1803 to 1805, subsisting on the local diet—mostly fish, shellfish and marine mammals—and becoming reasonably competent in Nootka language and ethnobiology. Though camas-root was not then produced by the Nootka, Jewitt notes that it was imported by them from their fellow Nootkan-speakers to the south, the Kla-iz-zarts (Classets, now Makahs), who called it quawnoose. He gives no Nootka word for camas but does cite cha-mass ‘fruit’ and its derivative cha-mas-sish ‘sweet’. Ethnologists knew of the importance of camas in Northwest Coast maritime commerce and of the presence of Nootkan vocabulary in CJ, and so adopted cha-mass as the etymon of English camas, the semantic connection being the sweet taste of fruit and of the baked camas-root. Once a Nootkan origin was accepted, it seemed to follow that the word must have found its way into English through CJ, as had high muckamuck and tyee. Unaware of the Sahaptian evidence—com-mas appeared in print in Patrick Gass’ Journal in 1807—the philologists had no workable alternative to Nootka cha-mass. But the first record of English camas, as remarked by Gary Moulton in his notes to the Journals of the Lewis & Clark Expedition, is clearly in a Nez Perce (NP) context. On September 20, 1805, Capt. William Clark identifies the etymon: “they [the Pierced Noses] gave us a Small piece of Buffalow meat, Some dried Salmon beries & roots..Some round and much like an onion which they call quamash”—and then introduces the word into English: “Emence quantity of the quawmash or Pas-shi-co root gathered and in piles about the plains”. (The influence of Shoshoni interpreters, including Sacagawea, is reflected in the frequent collocation in the journals of the NP word with the Shoshoni word for camas, pa:siko:, literally ‘water sego-lily’, so called presumably from its often moist habitat.) One couldn’t ask from Clark’s fickle and linguistically untutored spelling a better transcription of Nez Perce qém’es. (The alternation s ~ š may reflect a difference in NP dialects, as pointed out to me by Haruo Aoki, for whose patient help and magnificent Nez Perce Dictionary I thank him. Note the dialectically inclusive Linnæan name Camassia quamash!) Several points argue for a Sahaptian origin and against a Nootkan:

1.) Columbia River Sahaptin, a sister-language to NP in the Sahaptian family, has a cognate form, χmaaš ‘camas’ (from Eugene Hunn’s fascinating ethnography Nchí-Wána), which shows predictable correspondences between its consonants and those of the NP word, demonstrating a common Sahaptian heritage and making unlikely a recent borrowing from the lower Columbia River. And there is the striking analogy provided by a pair of Sahaptian words for another staple food-root, couse (cowse, cowish): NP qá:ws and Sahaptin χawš. (Again, NP is the source of the English word, whose earliest known record is in Lewis and Clark.)

2.) The Nootkan etymology requires an abrupt semantic shift from ‘berry’ to ‘camas’ (through ‘sweet’). Why should CJ have adopted for ‘camas’ the Nootka word for a fruit when Nootka had its own word—or one borrowed from Makah—for ‘camas’? (Doug Deur cites Hesquiaht Nootka kwanus, identical to Jewitt’s Kla-iz-zart quawnoose.)

3.) Clark’s earliest record is from the interior and is a closer match phonetically and semantically than the putative Nootkan etymon. Also suggesting an inland origin is the common CJ variant lakamas, which contains the prefixed French definite article common in Jargon words introduced from the east by French-speaking traders, and absent (mostly?) from words of maritime Indian origin. The several francophone members of the Lewis and Clark expedition were the first of hundreds to migrate west over the Rockies in the early 19th century. Chinook Jargon was at that time just beginning to expand up the lower Columbia River and was by the 1830s still largely unknown to speakers of Sahaptin and Nez Perce: missionaries’ sermons in CJ had to be interpreted into Sahaptian, and sign-language had to serve where interpreters weren’t available. The Jargon couldn’t have functioned as the intermediary for the introduction of a Nootka word into Sahaptian.

4.) The Nez Perces had no need of a Nootkan word for camas: their own country has been considered by native people to the west as the source of the best camas, and they doubtless had long had their own word for it. The blue-flowered plant of the lily family, whose starchy root was dug and pit-baked in late summer, was a staple throughout the Pacific Northwest. It was introduced to areas outside its natural range, especially in coastal prairies maintained by intentional burning, and was an important commodity in coastal trade, but it probably originated in the interior. Pointing even further north and east are the cognate Yakima Sahaptin and NP names for their Salishan neighbors, meaning literally ‘camas people’. And Lewis and Clark note that the camas of the coast and the Columbia valley had smaller roots and grew in smaller quantities than that of the “high rich flatts and vallees within the rocky mountains”.

In sum, the word camas was first borrowed into English and French from Nez Perce in 1805 and was probably introduced through English and French into the flourishing Chinook Jargon of the Northwest Coast and its hinterland. Thanks to Mike Cleven, Doug Deur, Sarah Thomason, Henry Zenk and other members of the active Chinook Jargon online community for their comments.