[CAUS-X]+N = N compounds have stayed fresh
Huh? Is this language or is it math!?
munk-hílu-pʰúli lamatsìn (image credit: McGill.ca)
I guess I’d better break down my headline’s formula, to show how incredibly interesting this structure in CW is.
- Phrases at least 3 morphemes long,
- Starting with (1) mamuk- ‘Causative’ plus (2) some root, modifying
- (3) a final Noun, which is the head of the phrase,
- Altogether taken as a compound (signaled by the “+”) noun phrase,
- Having a distinctive intonation pattern, such that the final Noun has secondary stress. (Stresses were not notated in most documents of CW, but from experience speaking it, we can infer itʹs different from mamuk-X followed by a N direct object, which has primary stress.)
- Semantically the difference is crystal-clear: the noun head of the phrases I’m talking about today is notionally the Active Subject of the Causative expression. (Mamuk-X + direct object expressions, i.e. non-compounds, obviously have that final Noun in a very different function from this.)
Highly noteworthy is the frequent occurrence of inanimate, even abstract, nouns as the head.
This shows a further degree of “grammaticalization” away from the literal meaning of mámuk ‘do; make’, and even beyond the Causative prefix mamuk-. (Which is limited to animate subjects in fluent CW.) We’ve arrived at a more general thing, a derivational affix.
This calls to mind early-creolized CW’s kəmtəks- “Characteristic” prefix (grammaticalized from kə́mtəks ‘to know’), which I’ve previously shown was also much applied to describe inanimate things — at its peak of usage anyways. However, a difference is that that prefix became apparently less often used from the late 1800s onward, at least partially reconnecting with its etymology, such that the kəmtəks- “Characteristic” expressions documented in the 2012 Grand Ronde dictionary all denote human propensities only.
(I’m also writing up an article about the negative-polarity equivalent of that, the wik-kəmtəks- “Characteristic Lack” expressions. Stay tuned.)
Back to our [CAUS-X]+N=N structure: when did it come into being?
There are zero instances of this structure in the earliest substantial document of CW usage, Demers-Blanchet-St Onge 1871 [1840s data].
But that could be chalked up to that book’s conciseness; its dictionary is limited mostly to listing root forms, hardly any compounds. And its textual contents are strictly in the service of teaching Indians how to worship the Christian God, so there’s hardly any discussion of the sort of earthly concerns that you’ll see listed among the examples below. Long story short, all occurrences of mamuk in that book have relatively literal meanings.
No [CAUS-X]+N=N examples are found, either, in the excellent Fort Vancouver-area data of Horatio Hale 1846. He too focuses on reporting simple root forms, largely because of an interest in showing the varied source languages of Jargon.
So far, I think there’s little reason either to assume [CAUS-X]+N=N existed, or that it was absent, in Fort Vancouver-era Chinuk Wawa.
But we do find plenty of these phrases in Father L-N St Onge’s 1892̂ manuscript dictionary, based on 1870s data.
Let’s discuss St Onge’s examples of this structure, including his sometimes-quaint translations into English. Speaking of stuff that looks like mathematical formulas, I’m using a linguist trick of bracketing the “constituents” that combine to form these expressions. I find that this strategy helps me see patterns within the general [CAUS-X]+N=N structure.
Let me stipulate that I’m purposely leaving out one particular type of expression, those involving tə́mtəm ‘heart, mind’; those have always felt like a complicated syntactic puzzle to me. (“Future research” department!) Thus, St Onge’s < mamuk-tlush-tomtom-wawa > ʹconsolationʹ is not in the following list.
I’d like to start with what might be called the simplest examples, 3-morpheme strings where mamuk-X directly modifies the head noun. That is, no direct object is expressed. (Caveat: sometimes in CW it’s hard to tell if a mamuk-X involves a conceptual direct object or not; keep reading.) There are lots of these:
ʹcurrierʹ (a worker in a stable; not a comb-maker)
‘brace, a support’
ʹdormitive; chloroform; anaesthetics; anodyne; narcotic; opiateʹ (sleep medicine)
ʹappetizerʹ (an appetite stimulant)
Here’s where I have to go into more detail about stress. Folks need to know that in Chinuk Wawa every “content-lexeme”, i.e. any morpheme that isn’t a grammar-function marker, has one, primary, stress. (Words that I write with more than one acute accent just have variable stress placement.) And according to my sense of the spoken language, every successive primary stress within a “syntactic word” (effectively just in compounds) is less strong than any that precede it. So then you wind up with what sounds like a secondary level of stress (I’ve shown it as e.g. à), a tertiary level (as with ȅ), and even a quaternary level (you’ll see one example below, and I’m not even gonna bother inventing a special stress mark for it!
I brought that up because the above type of compounds can themselves become part of a larger compound, leading to some interesting stress “downdrift” effects:
The head noun can itself be a compound:
And the “X” can be complex:
[[mamuk-míməlust] [kʰupa lúp]]+màn
I should point out that there are also 3-morpheme sequences where the middle one is more obviously the direct object, so these are like little twisted relative clauses, where the last word is semantically the Active Subject. So these examples below are pretty much synonymous with standard relative clauses: [mámuk mashín]+màn ‘machinist’ = mán mámuk mashín ‘man who makes machines’. These thus strike me as not being exactly the kind of compounds I’ve highlighted above, because these containing not that prefix mamuk- ‘CAUS’ but instead the literal (content-) word mámuk ‘do, make’:
ʹcapitolʹ (legislative building, place where laws are made)
- mámuk-maokamok-man [spelling SIC?]
ʹmachinistʹ (man who works on/with machines)
This type of compound can contain a compound direct object:
[mamuk [yáʔyim+[t’sə̀m-pȉpa]]]+man (man here would have quaternary stress 🙂 by my logic…hmmm!!)
Now back to the CAUSative expressions. One of these can itself take an overt direct object, all of this then modifying the head (notional Subject) noun:
ʹchisel; auger, bit’
ʹhartshornʹ (smelling salts)
The expressed object can be a compound:
ʹharvest moonʹ (month when growing things are harvested)
That last example is unusual, as its head noun ‘moon’ isn’t conceptually an Active Subject — I guess I see it more as a Stative, ‘to be the month…’
One more general observation about [mamuk-X]+N expressions is that the Noun head seems limited to a very small selection of generic roots (I’m defining this as ones that also head numerous other CW compounds):
- PRE-CONTACT ENTITIES
- mán ‘man’ (Note that it’s unremarkable for occupations to have been denoted by “man” compounds in the 1800s! Also note how CW distinguishes occupations from Characteristic traits, which are shown by the prefix kəmtəks-.)
- wáwa ‘talk, words’
- mámuk ‘action’
- háws ‘building’
- stík ‘wood’
- TYPICALLY POST-CONTACT ENTITIES
- múlá ‘machine / mill’
- pípa ‘writing’ (usually not ‘paper’!)
- chíkʰəmin ‘tool’ (not the word’s original meaning ‘metal’)
And a few less-generic roots (i.e. ones that occur in relatively few other CW compounds):
- PRE-CONTACT ENTITIES
- łúchmən ‘woman’
- lúp ‘rope’
- TYPICALLY POST-CONTACT ENTITIES
- lashén ‘chain’
- lamatsín ‘medicine’
- lashésh ‘chair’
Strikingly absent from that list is the most generic noun in the language, íkta ‘thing’, which occurs as the head of a large number of other compounds in CW. Perhaps that one root was exceptionally resistant to being treated as a conceptual Active Subject, we might guess — but then we’d have to deal with the fact that íkta mámuk ‘what makes/causes?’ is an old established way to ask ‘why’ in this language!
I don’t know how relevant my extra layer of taxonomy is up there, distinguishing Pre-Contact from Post-Contact root entities. But it’s safe to claim that the resulting compounds were all intended to refer to post-contact, European-framed, ideas. Hmm.
Historically, the [mamuk-X]+N forms tended to be replaced, by virtually the same expressions MINUS the mamuk-. This was especially true in northern CW, quite in keeping with that dialect’s trend of, first, simplification (as it was being rapidly introduced into areas such as British Columbia), then re-elaboration (using whatever resources the language had held onto).
But there are quite a few [CAUS-X]+N=N compounds nowadays at Grand Ronde! While my sense is that this structure ebbed somewhat between the Fort Vancouver era and the present day, it is definitely embraced now as part of the community’s linguistic heritage. While I don’t find examples of it in the 2012 GR Tribes dictionary, more recent learning materials show us phrases including these:
- munk-kə́mtəks tìlixam
- munk-hílu pʰúli lamatsìn
- munk-máksti pʰípa lemulà
‘copying or xerox machine’
(Grand Ronde writing style would involve writing each of these with more dashes. I’m leaving some out, to try showing their linguistic structure more clearly.)
PS: formatting so much linguistic data in WordPress is a lot of labour. This post took me hours and hours of work!