Taking a run at the etymology of “eulachon”
I’ve seen so many half-assed etymologies published for the English words that come from Pacific NW languages, I figure I’ll see if I can hold a candle(fish) to their absurdity.
Smelting, Cowlitz River, Longview, WA, February of 2016 (image credit: Columbia River Images)
Or more optimistically stated, let’s try to grease a trail to a better understanding.
Let me make it clear, I’m talking about the etymologies that appear in dictionaries of the English language; they seem not to have a habit of employing folks who have specialized knowledge when it comes to PNW languages and history.
“Eulachon” is a word variously synonymized as “candlefish”, “smelts”, “herring”, what have you. In any case it’s a fish of enormous cultural and economic importance.
The word is known throughout PNW coast English, with local pronunciations such as “hooligan”, “ooligan”, “oolikan”, and more. (Do you know another way to say it? Leave a comment below.)
We first know the word outside of its original purely Native context when Lewis and Clark’s expeditionary group note it in their journals on the lower Columbia River in 1805-1806. It was encountered in Chinookan territory, and it appears to be originally a Chinookan noun.
The 2012 Grand Ronde Tribes dictionary of CW lists úlx̣ən separately from the words of Grand Ronde elders, who perhaps had little experience with this fish once the people were removed from their traditional coastal lands and largely confined to inland reservations. Notice, if you will, that this pronunciation of the word has “plain” L, just as you might imagine from the common English spellings; this is Cowlitz tribal member Joe Peter’s pronunciation.
The other speaker cited in the úlx̣ən entry is Emma Luscier of Shoalwater Bay, Washington. She pronounced the word as iúlx̣ən, with an initial /i/. Now, because it’s assumed that this fish’s name is Chinookan, we infer that the u– is the usual feminine (singular) noun prefix, and that –lx̣ən is the noun stem that actually means ‘smelts’. Chinookan also has a prefix i- for masculine (singular) nouns, but it doesn’t allow a word to be both feminine and masculine at the same time; so GR 2012 says Emma’s pronunciation is a grammatical absurdity. To be fair, she was not a speaker of Chinookan, but she had plenty of experience of relatives speaking numerous languages of the region, so she added an i- to any word that she knew to be Chinookan.
I would modify GR 2012’s comment about Emma by pointing out that her iúlx̣ən is the single best documented match for Lewis & Clark’s spelling “eulachon”. So, in this case, maybe she wasn’t adding a spurious masculine i-. Maybe instead the root of the word starts with a vowel: –úlx̣ən. And maybe us linguists aren’t the only people ever to come up with a confusing analysis of this word:
“Eulachon” might then be just another of the countless Chinookan nouns that (sometimes) dropped their gender prefixes when they were taken into Chinuk Wawa. Thus we could wind up seeing variation between iúlx̣ən and úlx̣ən, both pronunciations seemingly documented by the different English spellings of the word, and thus both apparently legitimate CW.
(Not central to today’s quest, but worthy of noting, is that whether this noun stem has u- or i- on it in Chinookan, it’s marked as a singular, even though you’d never just catch one of this small fish, and you’d never trade just one. Perhaps they were seen as a mass rather than as individuals.)
In GR 2012’s entry for the word, I see that by the time of publication they had located one clear Chinookan cognate glossed as ‘smelts’, which I’ll phonologize as íɬx̣ən; it’s found in Edward R. Curtis’s “The American Indian”, volume 8. Note that that form has the “slurpy” /ɬ/ sound, not “plain” /l/. (We don’t know why this change occurred; I can only tell you that exactly this pair of sounds sometimes varies unpredictably in regional languages such as the SW Washington Salish ones.)
Curtis’s form implies the consonant-initial root option, as does GR 2012’s closest match for their CW headword, ú-ɬx̣an ‘dried eulachon’ from an unspecified Chinookan language.
GR 2012 make the valid point that possibly the same stem in Chinookan means the fish (generically?) when it has the masculine prefix i-, and the dried fish (an item of trade, thus more likely to come into Chinuk Wawa) when it has the feminine u-. They don’t say in so many words, but it’s well-known that Chinookan languages use their gender prefixes for additional purposes, such as feminine-marking to create diminutives. And I expect a dried smelt is smaller than a live one!
I’d like to contribute some further detailed data on the Chinookan source of “eulachon”. I’m going to start with the farthest upstream language in that family, working downriver.
- UPPER CHINOOKAN:
- Kiksht (Wasco, Wishram, Warm Springs, etc.)
—– (no forms detected)
i-ɬx̣ə́na ‘smelts’ (masculine; on racks, in the process of being smoke-dried)
- Kiksht (Wasco, Wishram, Warm Springs, etc.)
- LOWER CHINOOKAN:
i-ɬx̣ə́na ‘smelts’ (masculine; generic reference and the live fish)
a-ɬx̣ə́na ‘smelts’ (in Kathlamet a- is the usual feminine singular prefix) (in the Kathlamet Texts, this form is used when someone is roasting a limited quantity for a person’s meal)
- Natítanui (Shoalwater, Clatsop)
í-ɬx̣ən ‘smelt’ (masculine; generic, live, as specifier of the kind of guardian spirit a person has [spirit songs are associated with smelts also in Kathlamet])
ú-ɬx̣ən ‘a smelt’ (feminine; in this case, one being hung up to dry)
So, there are some patterns there.
- We don’t find smelt words way up the Columbia, and I reckon that’s partly because smelt don’t run that far from the ocean.
- Clackamas Upper Chinookan & Kathlamet Lower Chinookan, which neighbor each other, share a stressed stem form, -ɬx̣ə́na, versus the distinct unstressed -ɬx̣ən of Natítanui, the form and prosody of which best match the known occurrences of the word in Chinuk Wawa and English.
- The available evidence confirms the idea that Chinookan masculine prefixes indicate this fish generically, whereas feminine marking is associated with its use as a product.
- In this regard it’s once again Natítanui, with its u-, that most closely matches the first vowel of known form(s) of the word in Chinook Jargon.
- That form correlates with the use of smelts as a product (food; trade item), which in turn fits the known picture of Lower Chinookan trading contact with newcomers which apparently gave rise to the pidgin-creole Chinuk Wawa.
- We find no Chinookan-internal evidence for Emma Luscier’s iúlx̣ən.
Further evidence, to my way of thinking, to show that “eulachon” is natively a Chinookan word and an item of trade comes from its form in the neighboring and almost intertwined ɬew’ál’məš (Lower Chehalis Salish) — ʔúɬx̣. This is because it makes sense within Lower Chehalis to have perceived a Chinookan word úɬx̣ən as containing the Salish suffix -(ə)n(‘) ‘Instrument; tool/thing for’, which speakers used on various words for trade items, almost as if it were the analog to Chinookan u-. Even though this entails an otherwise unknown Salish root ʔúɬx̣, that root indeed became the Lower Chehalis word for ‘smelts’!
As far as we can tell, it wouldn’t have occurred to Chinookan speakers to have added an otherwise nonexistent suffix (in their language) -(ə)n(‘) to a hypothetical foreign noun úɬx̣.
It goes to show you that languages are way more comfortable imagining new roots than new affixes…
(Image credit: Northwest Coast Archaeology)
The eulachon has been famously called the “salvation fish”, and folks claim this is because its run comes at the end of winter when traditionally the people’s stores of food were running out. It happens that a SW Washington Salish root for ‘save, take care of’ resembles “eulachon”, although it contains the “plain L” sound; an example is Lower Cowlitz ʔúlx̣-n (which means approximately ‘(s)he takes care of’ someone, with a different -n suffix). I don’t see traces of this root in the known data on Lower Chehalis or Quinault, but the same verb form in Upper Chehalis is additionally used for ‘gather (berries)’; could it have applied to smelt-fishing in more than one sense? Anyhow, the Cowlitz River appears to be the longest, most productive smelt run off the main Columbia River, and it’s conceivable that folks around there folk-etymologized the Chinookan word into a Salish one, maybe making a Salish pun like the many others I’ve noticed in these languages. (Their native Salish words for “eulachon” are totally different, the most common one being (s-)qʷálə́sti(ʔ).)
And that might account for why this word in Chinuk Wawa has the “plain L” sound.