Advances in Role and Reference Grammar

(singke) #1
CLAUSE COMBINING IN NOOTKA 267

tion is a kind of compounding. Anderson (1985a:53-54) also makes this distinc­
tion, referring to Northern Wakashan Kwakw'ala (= Kwakiutl). The lexical suf­
fixes of Nootka, of which several types must be distinguished (including these suf­
fixal verbs), do not themselves involve productive incorporation. But certainly
these patterns hark back to ancient compounds. Swadesh (1948) isolates a late
layer of lexical suffixes that can be equated with stems. See Mithun 1984b:887-889
for a sampling of Nootka types (where (172) and (178) illustrate the suffixal
verbs); but note that there is not just an "etymological relationship" between
  and suffix -nak "have"; rather the former contains this suffix (basic form
-nakw) added to the referential stem 2u- (the variable-length vowel of the suffix
is short in a third or later syllable).


  1. In the text sample considered below, for main clauses in narrative passages, the
    types SV and SVO together make up only 2.3% of the occurrences, so that 97.7%
    of the clauses are verb-initial. In order of declining frequency the proportions are:
    V 42.6%, VS 31.2%, VO 18.0%, VSO 3.6%, VOS 0.7%, and other verb-initial
    1.6%. Clauses with two or more arguments (VSO, VOS, and others) make up
    only 5.6% of the occurrences. In quotations one finds in main clauses: V 78.8%,
    VO 12.1%, VS 8.1%, and other verb-initial 1.0%. Rose (1981:ii, 192, 194) also
    describes (Kyuquot) Nootka as VSO, and she gives (179-182) comparable statis­
    tics, showing an even smaller proportion of clauses with more than one adjunct,
    along with discussion. Similarly,Whistler (1985:234): "the preferred Nootka order
    of arguments is subject-object, but ... that is not an ironclad rule." On the other
    hand, Sapir (1924:83, η. 4) states: "verb, object, subject — this is the most com­
    mon Nootka order." In the text he is analyzing there are several occurrences of
    VOS clauses near the beginning, but overall the VSO type slightly predominates:
    the proportions are respectively 6.2% and 8.0% of main clauses in narrative. The
    proportion of clauses with two arguments is much higher in this text: 18.6%.

  2. Nootka would belong to Greenberg's (1963:87) and Hawkins's (1983:283) Type 2,
    which is characterized by verb-initial word order, prepositions rather than post­
    positions, and possessed-possessor and adjective-noun orders. In the same type,
    in Hawkins's Expanded Sample, with specifically VSO word order, are found
    Northern Wakashan Kwakiutl and Chimakuan Quileute, which borders Nootkan
    Makah on the south. The other included northwest coast language, Coast Salish
    Squamish, has less specific verb-initial word order. This type constitutes an excep­
    tion to Greenberg's (1963:67, 89) statistical generalization (Universal 17): "With
    overwhelmingly more than chance frequency, languages with dominant order
    VSO have the adjective after the noun." The reasons for this exception hinge on
    the fact that these word-order studies have defined their parts of speech, in par­
    ticular adjective, notionally rather than on distributional grounds for the indi­
    vidual languages. Adjectives in Nootka are a subtype of verb, so the language is
    merely being consistently verb-initial when they come before their nouns. (Much
    the same point is made by Rose 1981:194). Similarly for the other stigmata of this
    type. The prepositions are also a subclass of verb. Even the possessed nouns (with
    third person ending -uk-2i) are nominalized possessive predications. Likewise for
    another Nootka phrasal pattern: numeral-noun.

Free download pdf