On the size of the Chinuk Wawa “Adjective” class


Here’s an adjective falsely rumored to come from “Chinook” (image credit: Curious.com)

It’s rather small.

That’s typical for a language in this world.

It’s probably shockingly small, from the perspective of say an English- or French-speaker. You can make a solid case that those Northwest European languages contain many hundreds, probably thousands, of Adjectives. (Using the definitions you’ll see below.)

Chinook Jargon’s Adjective class is 1 or 2 “orders of magnitude” smaller than that — I mean, something like 10 or 100 times smaller!

Defining “Adjective” for our present purpose: inspired by linguist RMW Dixon, I’ll say these are whichever root-forms in a language have a main purpose of describing rather than identifying entities, and can’t be assigned to some other major class such as Verbs or Nouns.

Additional housework:

I’m going to count only “simplex” forms in today’s study; the useful assumption in this track of research is that we’re thinking only about “basic” Adjectives, and not the many that are “derived” by combining simplex elements (historical linguistics research tells us that these usually are newer, less fundamental items in the lexicon).

Also, I’ll try to focus on those words that are clearly primarily Adjectival in meaning and use. There are quite a few words in a pidgin-creole like the Jargon that are “multi-functional”, showing up in multiple uses (e.g. as both an Adjective and an Adverb), with a non-Adjectival use being predominant, and I feel those should be flagged. So I’ll enclose those in parentheses below.

And I’m going to stick to a relatively coherent data set, so I’ll limit my view to the recent elders’ Grand Ronde Reservation variety of Chinuk Wawa. Other, more northerly dialects will have somewhat different inventories of Adjectives, but, I predict, the same overall level of complexity in this regard. Ditto for older varieties. In both cases there seem to have been a somewhat greater, but comparable, number of Adjectives than in our present data set.


Evaluating the types of Adjectives: to refer to RMW Dixon again, he summarizes what linguists have discovered, showing that 4 types of Adjective seem to be present in all human languages.

(From here on in, I’m referring to Dixon’s summaries in the 2004 book “Adjective Classes: A Cross-Linguistic Typology” and his 2010 book “Basic Linguistic Theory”.)

These Types 1-4 thus form the minimal repertory in any language. Here’s the Chinuk Wawa rundown:

    • háyásh ‘big’
    • tunús ‘little’
    • yútskat ‘short’
    • yúłqat ‘long’
    • łə́q’əł ‘wide’
    • p’íłił ‘thick’
  • 2 AGE:
    • úl ‘old’
    • (chxí) ‘new; recent’
    • ((ánqati)) ‘ancient, from long ago’
  • 3 COLOUR: By the way, there’s an enormous research literature on this class, showing exactly which colour terms a language will most likely have, depending on the number of them that it has. The “Berlin & Kay hierarchy” (here copied from page 213 of Greville Corbett’s chapter in the 2004 book I’ve mentioned) predicts the following range from most frequently occurring in languages to least frequent: [WHITE, BLACK] < [RED] < [GREEN, YELLOW] < [BLUE] < [BROWN] < [PURPLE, PINK, ORANGE, GREY]. You’ll see that Chinuk Wawa fits this hierarchy well:
    • łíʔil ‘black’
    • tk’úp ‘white’
    • pʰíl ‘red’ (evidently all warm colors, originally, as it’s used in words for ‘gold’ and so on)
    • pchíx̣ ‘green <=> blue’ (the ‘grue’ of linguistic anthropology)
    • (spúʔuq) ‘faded; [by extension] grey’
    • (t’sə́m) ‘marked, spotted’
    • (words for several more colors don’t count, as they refer only to horses: líkʰrém ‘cream-colored’, sándəli ‘roan-color’, etc.)
  • 4 VALUE:
    • łúsh ‘good’
    • t’úkti ‘pretty, handsome’
    • kʰə́ltəs ‘no-good, bad, worthless’ [and ‘old, worn-out’ when ref. inanimate things]
    • x̣lúyma ‘different, strange, weird’
    • (pʰishák) ‘wild, unruly, bad’
    • (drét) ‘real’


Further classes of adjectives may show up, in “medium-sized and large adjective classes” (2004:4).

I’ll note here that Chinook Jargon has quite a number of ethnicity descriptors that can be used as modifiers of Nouns, though virtually all of them are fundamentally Nouns themselves: sháwásh, bástən, chinúk, etc. Their Adjective-like use is a byproduct of this language’s grammar of possession, which expresses an innate / inalienable property of an entity as a Noun-Noun compound: sháwásh-íliʔi (literally ‘Indian-land’, for ‘reservation’) and chinúk-wáwa (‘Chinook-speech’) are thus the same phenomenon as kʰíyutən-úyxat (‘horse trail’) and lipúm-tsə́qw (‘apple-liquid’ for ‘cider’).

Adjective type 5, in languages with just more than the minimum Adj inventory:

  • 5 PHYSICAL PROPERTY [other than Dimension, above]; apparently the largest class of CW Adjectives, perhaps having to do with the Chinookan source languages’s heavy use of descriptive ideophones / onomatopoeias:
    • tʰíl ‘heavy’
    • q’ə́l ‘hard’
    • t’łímin ‘soft, mashed’
    • łə́q’əł ‘flat’
    • lúʔluʔ ’round’
    • skúkum ‘tough, strong’
    • q’áyʔwa ‘crooked’
    • wám ‘hot, warm’
    • kʰúl ‘cold’
    • tláy ‘dry’
    • t’sí ‘sweet’
    • pʰúli ‘rotten’
    • chxə́p ‘extinguished, “out” ‘
    • ch’úx̣ ‘peeled, skinned; scraped’
    • stúx ‘untied’
    • k’áw ‘tied up’
    • íx̣puy ‘covered, closed, shut’
    • x̣álaqł ‘open’
    • láx̣w ‘tipped, lopsided, leaning’
    • łúk ‘broken’
    • pʰáł ‘full’
    • q’wetł ‘tight’
    • sháp ‘sharp’
    • t’wáx̣ ‘bright, light, fiery’
    • t’łə́x̣ ‘torn, ripped’
    • t’łíł ‘bitter’
    • t’sə́x̣ ‘split’
    • x̣ə́nłq’i ‘crooked, bent’
    • (ípsət) ‘hidden’
    • (hə́m) ‘smelly, stinky’
    • (drét) ‘straight’
    • (páya) ‘ripe; cooked’
    • (sáx̣ali) ‘high’
    • (kíkwəli) ‘low’
    • (sáyá) ‘far away’
    • “And a sub-class referring to corporeal properties, e.g. ‘well’, ‘sick’, ‘tired’, ‘dead’, ‘absent’ — 2004:4)
      • míməlust ‘dead’
      • sík ‘sick’

Adjective type 6, only likely to occur in medium-to-large inventories:

  • 6 HUMAN PROPENSITY (such as ‘jealous’, ‘happy’, ‘kind’, ‘clever’, ‘generous’, ‘cruel’, ‘proud’, ‘ashamed’, ‘eager’, cf. 2004:4) — relatively few in CW, and in line with recent observations of mine, there’s a remarkable tendency for Jargon’s descriptors to carry a negative overtone:
    • yútłił ‘proud, arrogant; glad, happy’
    • píltən ‘crazy; foolish’
    • klísi ‘crazy’
    • lísi ‘lazy’
    • łax̣áyam ‘poor, pitiful, wretched’
    • łík ‘silly’
    • łíx ‘sexually aroused, horny’
    • másháchi ‘evil, bad’
    • sáliks ‘angry’
    • shím ‘ashamed, embarrassed’
    • yíx ‘tipsy; deluded’
    • (k’wás) is often translated as ‘afraid’ but all evidence suggests to me it’s a verb ‘to fear’.
      As Dixon has pointed out, in languages having smallish Adjective inventories, HUMAN PROPENSITY concepts are likely to be expressed in other ways, and sure enough CW usually has X-tə́mtəm (‘-heart’) and X-latét (‘-head’) phrases for emotional and intellectual states, plus kakwa-X (‘like-X’) for describing both animate and inanimate entities (as opposed, interestingly, to X-kakwa for non-animates only!). Relevant to this point, Dixon reports on a study that found kids growing up learning English tend at first to use the PHYSICAL PROPERTY & HUMAN PROPENSITY terms predicatively (like in ‘Emma is nice’), unlike types 1-4 (plus SPEED), which they tend to use attributively (like in ‘good doggie’).

Adjective type 7, apparently in yet larger inventories (its concepts are often expressed as verbs etc. in other languages) :

  • 7 SPEED: none; this is expressed in CW by Adverbs, not by Adjectives: (áyáq) ‘fast, rapid’ (łáwá) ‘slow’.


In truly large Adjective classes, languages may have these additional types (the English examples here are from 2004:5) :

  • 8 DIFFICULTY (‘easy’, ‘difficult’, ‘tough’, ‘hard’, ‘simple’, etc.) —
    CW expresses these concepts by derivations from Extended Class 5. 
  • 9 SIMILARITY (‘like’, ‘unlike’, ‘similar’, ‘different(/strange)’, ‘other’, etc.) —
    CW expresses these concepts by derivations from Core Class 4 & from adverb/conjunction ‘kakwa’ & from Class 13 the Number ‘1’.
  • 10 QUALIFICATION (‘definite’, ‘true’, ‘probable’, ‘possible’, ‘likely’, ‘usual’, ‘normal’, ‘common’, ‘correct’, ‘appropriate’, ‘sensible’, etc.) —
    CW expresses these concepts by derivations from Core Class 4 & by using discourse markers.
  • 11 QUANTIFICATION (‘all(/whole’), ‘many’, ‘some’, ‘few’, ‘only’, ‘enough’, etc.) —
    In CW such roots are better analyzed as Quantifiers, together with Numbers.
  • 12 POSITION (‘high’, ‘low’, ‘near’, ‘far/distant’, ‘right’, ‘left(/strange)’, ‘northern’, etc.) —
    In CW these concepts are nearly always expressed with Adverbs.
  • 13 NUMBERS (‘1’, ‘2’, ‘3’, ‘4’, etc.; ‘first’, ‘last’, other ordinal numbers) —
    CW treats these and other Quantifiers distinctly from Adjectives per se.

Summarizing what is probably a long-feeling, dense post for my non-linguist readers (sorry!!):

Chinuk Wawa has a medium-size set of Adjective roots. It only has the expected 4 Core classes plus the then-predicted Class 5 ‘Physical Properties’ and some of Class 6 ‘Human Propensity’.

This may not be very shocking, as CW’s total vocabulary of roots is fairly slim. 

Besides, being (as far as the written record can tell us) a relatively new language, we might expect CW to give just such a result in our investigation of its Adjective lexicon. In other words, it hasn’t had much time to expand its collection of words.

On those grounds, it would be fair to be surprised that CW has this many Adjectives!

However, what also should be done is to compare Chinuk Wawa’s Adjective class structure, not only with those of English and French (which we know to be of the “very large” type), but with those of the Indigenous source languages (at minimum Chinookan and Salish). It may be that those Native languages are like the thousands of other languages that have few Adj’s.

I will try to do that kind of comparison, at least with the important Lower Chehalis Salish & Shoalwater-Clatsop Lower Chinookan as representative samples, when I come to the point of writing up this research for publication.

And I would very much like to see further research of this kind on the Adjective classes of other pidgin, creole, and mixed languages. Such “contact languages” are typically overlooked in professional linguists’ work on language typology & universals.

Finally I want to observe that we can see a lot of Chinuk Wawa’s history when we look through its Adjective inventory. “Linguistic archaeology”, as I’m known for calling this kind of historical linguistics, is inseparable from any serious research into a language’s real-world use. Above, for instance, we’ve seen that quite a lot of ideas we might expect to see put as Adjectives in a language actually trace, demonstrably, to other kinds of words that came earlier in CW’s development.

What do you think?