Some Northern Hokan Plant-Tree-Bush Forms
Total Page:16
File Type:pdf, Size:1020Kb
UC Merced The Journal of California Anthropology Title Some Northern Hokan Plant-Tree-Bush Forms Permalink https://escholarship.org/uc/item/7494t408 Journal The Journal of California Anthropology, 1(1) Author Silver, Shirley Publication Date 1974-07-01 Peer reviewed eScholarship.org Powered by the California Digital Library University of California Some Northern Hokan Plant-Tree-Bnsh Forms* SHIRLEY SILVER N the conclusion of her paper on Cali lects, and Kashaya Pomo (Pk), and function I fornia Hokan, Mary Haas (1964) com either as members of noun phrasal or stem ments on intersection, an important problem compounds or as substantival suffixes. in Califomia linguistic prehistory. In compar The constituents of the SH, AC, AT, and ing some Proto-Hokan body-part terms with CH terms are noun stem + suffix. The K and Algon-Ritwan, Yukian, and Penutian similars, Y constituents are noun stem + noun stem. she demonstrates likely intersections between The K morpheme occurs only as a postpound Hokan and Algon-Ritwan, Hokan and Yukian, except in the noun theme Hpaha 'tree'. In Y and Hokan and Penutian. Haas (personal we also find the substantives '^i(na) in Central communication) uses the term intersection to and Northem Yana and ?/w/ m Yahi, with the call attention to similarities between forms (in meanings 'stick, wood, tree' (-na is an absolu- presumably genetically unrelated languages) tive suffix and -wi is probably the plural for which coincidence seems an unlikely marker [with cohective impHcation]). explanation and for which no evidence as yet The Pk term for 'sugar pine' is a type of exists to support an adequate genetic or noun phrase; however, Pk also uses noun stem diffusidnal explanation. compounding, e.g., "^ihte 'feather' + q^ale In this paper, I extend into the domain of 'plant' > t^eq^^ale 'elderberry bush' (Oswalt, plant life the consideration of intersection personal communication). Pk q^ale is a reflex begun by Haas within the framework of of Proto-Pomo (PPo) *qha • le 'tree', a form I body-part terms. Under discussion here are the diachronic complexities manifested by some northern Hokan morphemes occurring in terms for plants, trees, and bushes. These SUGAR PINE TERMS morphemes, all having a general interpretation Sugar Pine Nut Sugar Pine 'of the plant world', make up two separate phonological subsets whose members are cog SH ac'aw acawhihii nate. The fact that there are extra-Hokan AC assa assalo similars for each subset leads to consideration AT ac'^aw ac''awu:p of the possible extra-Hokan connections. CH hakew hakewna'a The morphemes in question occur in the o o K ''u's terms for 'sugar pine' (see Table 1) in Shasta ''usip (SH), Achumawi (AC), Atsugewi (AT), Chi Y cau(na) cau'i mariko (CH), Karok (K), the Yana (Y) dia Pk cuye' cuye' q'lale PLANT-TREE-BUSH FORMS 103 assume to have two constituents, *q'^a() + that *({^a • le had become synchronically un- *le. The basis for this assumption is as segmentable, i.e., it had undergone morphemi- follows. Robert Oswalt (1964), in a paper on zation.^ In reconstructing PPo *-6', an ele Pomo intemal relationships, cites PPo *k^ah- ment occurring in PPo forms for 'cotton- wa • (I), later revised to *q^ahwa • (I) 'bark (of wood', 'maple', 'live oak', 'pepperwood tree', a tree)', a reconstmction that fits every Pomo and 'soaproot', McLendon (1973:49) presents language except Northem Pomo (Pn); i.e., the other evidence for morphemization in PPo second syllable of Pn ^ewa 'bark' fits well plant-tree-bush forms in pointing out that the with the second syUable of *q^ahwa • (I), reflexes of *-b'^ are segmentable only in the whereas the first syUable does not fohow Pk, Pe (Eastem Pomo), Pn, and Pse forms for established sound correspondences. Oswalt's 'pepperwood tree' (cp. PPo *bah • e 'pepper- suggestion that sewa may be a blend of two wood nut', *bah eb'^ 'pepperwood tree'). elements, the second syllable representing a She also suggests that PPo *-b'^ involves cognate element and the first syllable a metathesis of an older Pre-Pomo **-^b (result non-cognate element, is supported by Sally ing from deletion of a Proto-Hokan */). McLendon's (1973:66) discussion of the aber To sum up at this point, there are two rant Pn form as a reflection of a proto form PPo elements, *-le and *-b''', both of which *k^e(h)wa or *xe(h)wa, or an early borrow occur in PPo forms for plants, trees, and ing of a similarly shaped form from Central or bushes; *-le appears to have been nonproduc Southeastem Pomo (Pc, Pse). The cognateness tive and *-b^ may have been quasi-productive. of the second syhable of sewa is an argument These two elements and the other northern for the morphemic complexity of *q^^a- Hokan plant-tree-bush markers shown in Ta hwa • (I), and comparing *q^a • le and *ciha- ble 1 fall into the two phonological sets given hwa • (I) raises the possibility that the first ki Table 2. syhables of these forms are variant shapes of In Table 2 both sets contain PPo and AC the same morpheme. If this is the case, forms. As is the case with PPo *-le, the *q^a • le is also morphemicahy complex and evidence for AC -Id • is tenuous. In his *-le represents an element that may be sec dictionary of Achumawi, Olmsted (1966) lists ond member of a noun stem compound. -Id • 'tree (suffix to name of fruit)' in the It is worth noting here that the remarks Achumawi-English section under L. In the Sapir (1917:17) made about Yana secondary section under 0 the entry -o • (-Id • after stems may weh hold not only for other areas of the Yana grammatical system but also for Table 2 other areas of the grammatical systems of PHONOLOGICAL SETS other Hokan languages, including the Pomo group: "... it is highly probable that they Set I Set 11 were originally primary stems which, being AC 0- 1 6- often compounded with other preceding primary stems, gradually came to lose their CH n (a)'> a independence ...[;] ... loss of independence PPo '' b Ob'') 1 e of secondary verb stems has evidently been SH h i- h u consummated as a general process in the very AT u: p remote past." I assume here that *-/e' had lost its K C) i- P independence by the time of PPo and also Y •^ i 104 THE JOURNAL OF CALIFORNIA ANTHROPOLOGY vowel) 'tree' implies that there is one mor (set II) occur in apparent contrastive distribu pheme with two phonological shapes; how tion, -d • might be explainable as the result of ever, in his work on Proto-Palaihnihan (PPa) diffusion from AT; however, diffusion is a phonology, Olmsted (1964:67) compares AC more likely explanation for AC -u • p in -6 ~ -o • with AT -o • p and reconstmcts PPa pa'^yipu • p 'mountain mahogany' (cf. AT *o • p 'suffix denoting plant'. Since this re winayu:p 'mountain mahogany'). (Leonard constmction does not account for the variant Talmy [personal communication] says that shape -Id •, one imphcation is that AC -Id • -u:p [the AT entry in set I] is the probable and -6 • are two morphemes with same or underlying shape; colon represents a phono near-same meanings rather than phonological logical segment whose phonetic effect, under altemants. For further support of the two- as yet incompletely specified conditions, morpheme argument compare AC slao • 'tam lengthens and lowers a preceding vowel.) arack tree {Larix laricinaY, {sla 'tamarack K provides us with another possible case nut'), and slald 'Murray pine {Pinus balfouri- of diffusion. Haas (1964) suggests that K ana)'(Olmsted 1966). '^dhup 'wood' is a formation of ^a • h 'fire' + It is certainly feasible to consider the ('^)i-p, with -up as a rare allomorph of members of set I as cognate. A tentative PH (^)i-p^ (see also kusup 'plant, sp'. Bright reconstmction *h/^i • p^u, closely paraheling 1957:364). -up could have come into K from Proto-Shastan *hi • p^u (SH hi • hti; Konomi- one of the neighboring Shastan languages, hu hip(h)u as in kuwakinahip(h)u'^ 'peppemut e.g., New River Shasta. bush' (kuwakhin 'peppemut'); Okwanuchu AC -u • p and K -up are obviously part of -ihu as in atsahihu 'sugar pine') would imply: the diachronic problem exemplified by the (1) that in the Palaihnihan languages AC and forms in set I; whether they are instances of AT (a) the vowel of the first syllable assimilat borrowing or are to be explained in terms of ed to the vowel of the second, (b) the second historical phonological processes must be de vowel was lost, and (c) in AC the consequent termined through intemal reconstruction. final consonant was deleted; (2) that the Although the tentative PH reconstmction second vowel was lost in K and Y, and Y also for set I is no more than a 'housekeeping' deleted a consequent final consonant. form and it is uncertain whether the members We cannot mle out the possibility that the of set I are reflexes of proto-bimorphemic Y form, at least, may be a reflex of only one element, there can be Httle doubt that the set member of a proto-bimorphemic string, espe- I forms are cognate, not only with each other ciahy since we do not have for Hokan an at but also with forms from Yuman languages in tested body of phonological rules which spe southem and Baja Cahfomia. In Yuma ?/• cify the conditions under which final conso occurs as a first position element in noun nants would be deleted.