! f{lA Tiny fCt"h. The Mu.Jinypata Language of North-West Australia by Michael James Walsh I I. , .. . /~.#l, A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy at the Australian National University. September 1976 I \ -:-----, r?~ ~;;;;7 -Ahb~ev1ations and Conventions. L·";;i=-j,-~,_ ,~;gie-- i .1} pc pauea! number (from three to around ten) ABS "~ absolutive case .[1 ?L ho~ W"I-l- 1"1i-b :w p1 piural number (more than paucal'ie. about ten or more) Aux auxiliary j RECIP reciprocal BEN I benefact1ve REDUP reduplicated COM com:lCative ~i REFL reflexive DAT dative ~ it. lIl!. redundancy rule du dual number 'f> SBa syllable boundary rule DUB dubitacive ss singular number EMPH emphatic SlB "sibling". Used to refer to a group whose members ERG ergative case are of the same subsection exc exclusive (pronoun) SR stress rule FEM feminine (at least one member of the group is female) TOPIC topicalizer FlIT future tense VR verb root HAD habituative/habltual .~ II!PEBF imperfective aspect if!; 1 first person inc inc1usive (pronoun) ~. 2 second person INST ·instrumental case 3 third person INTEBROG interrogative- LOC locative Person-number "statusll combinations for pronouns appear as HASC masculine (the group consists entirely of males) 2.g MP morphphonological rule. A particular rule is 1 du exc MASC identified by a number ego MP-6 [2.1.10.17]. Ne 3 pc SI:B noun classifier NEG negative ) indicates optionality NI number indicator NP noun phrase .~ indicates ~hat at least one of the enclosed must be c.hosen PERF perfective aspect .; ',! \ • when appearing before a form or sentence indicates The farms are given in underlying form in the morpheme-by unacceptability for native speakers (whether &. morpheme present~tion since it is not possible to non-arbitrarily grammaticality or semantic anomaly) or otherwise aSSign morpheme cuts to a form after the morphophonological indicates an underlying form to which a morphophonological rules have acted on the form. rule must he applied before it attains an acceptable surface form. syllable boundary When appearing after some word in a text it refers to a footnote at the bottom of that page. Mu~inypata sentences are generally supplied with an interlinear gloss, morpheme-hy-morpheme. together with an English translation of the complete sentence. Non-pronominal NP1s do not specify number or def10itenesa while pronominal NP's are particularly detailed in their reference: in the interests of readability the former tend to be overspecified in the English translation while the latter are underspeci£ied: 'the/a old man/men/bo8s(es) is/are sitting over there (remote)' I might be rendered as "the old man is Sitting over there" t (Mu~inyp~ta : pule pa~u) and "we two exclusive non-siblings at least one of whom Is female will wash the lower legs of you few (three to about ten) ,siblings" which represents a fairly full translation of the Mu~inypata~: manantarmuPulnu~lnda is simplIfied in Its translation. Content.s. Word' Classes 125 .1 Introduction 3. 1 Nominal 125 1. 1 Linguis tic Type 1 3.1.1 Pronoun 127 1.2 Dialect Situation 4 3.1.2 Noun-classifier 127 1.2.4.A Language Map of Port Kests Area 5 3.1.3 Adjec~ive 127 1.3 Surrounding Languages 7 3.1.4 Noun 129 1.4 Language Names 9 3.2 O~her word classes 129 1.5 Background 15 3.2.1 Adverb 129 1.6 Recent History of the Area 16 3.2.2 Adjective/Adverb Qualifier 129 1.7 Previous Work 19 3.3 Criteria for Identification of a word 133 ffr.'fJ4 3.3.1 Mobility 133 1\"1. 2 Phonology .~~ 3.3.2 Isolabability 133 =;~:': 2.1 Segmenta1 Phonology 24 3.3.3 Pausing 134 2.1.1 Inventory 24 3.4 Verb 134 2.1.2 Residual problems in the Inventory 25 •. 3.5 Int.erjection 134 -;- .;:j;! 2.1. 3 Comparison with other Australian language~ 26 .. ', .. 3.6 Semantic Content of the Open Word Classes 137 'a" ··.'.···,·.'.. 2.1.4 Phonemic Contrasts 27 .:;, 3.6.1 Nouns 137 2.1.5 Realisations of Consonant Phonemes 31 ' 3.6.2 Adjectives 138 2.1.6 Distribution of Phonemes 34 3.6.3 Verbs 138 2.1. 7 Vowels 45 3.1 Particles 139 2.1.8 Discussion of Interpretation of Stop Phonemes 49 3.8 Shared Characteristics of the Word Classes 139 " 2.1.9 Distinctive Feature System 53 3.9 NOml-C1assification 141 2.1.10 Marphophonemic Change 69 3.9.1 Inventory 141 2.2 Suprasegmental Phonology 104 3.9.2 Discussion 143 2.2.1 Syllables 104 J.9.2.B Map of Daly F~ly Languages 149 2.2.2 Stress 106 -Morphology 150 4. 1 Morphology of Nominals 150 4.1.1 Pronominal Morphology 150 4.1.2 Case Inflections 163 \ 4.3.6 nukun 247 4.1.3 Demonstratives 173 4.3.7 'lata 247 4.1.4 Body Part Incorporation 175 4.3.8' ylcja 248 - in nominala 175 4.3.9 ejlyiga 248 - derivational affix, -ma 183 4.3.10 wa~a 248 4.1.5 IDS.- 192 4.3.11 man! 250 4. I. 6 Number Systems 194 4.3.12 yuwu 250 4.1.7 Special Non-Singulars 199 4.3.13 l)inlpunV 251 4.1.8 Reduplication 201 4.3.14 katu 251 4.2 Verbal MOrphology 202 4.3.15 W81)U 252 4.2. I General Statement -on Verb Complex 202 4.3.16 l)a.J8 252 4.2.2 General Statement on Auxiliaries 212 4.3.17 kal)gal 253 4.2.3 Tense and Mood 213 4.3.18 pepe 254 4.2.4 Negation in Verbs 215 4.4 Interrogative Words 254 4.2.5 Analysis of Auxiliaries 218 4.4.1 l)a.J8 254 4.2.6 ~ersanal Verbs 231 4.4.2 taoku 256 4.2.7 Middle Verbs 234 4.4.3 na!)kat 257 4.2.8 Reflerlves 234 4.4.4 m!nYdYI.J6 259 4.2.9 ReciprQcal~ 236 4.4.5 nan 260 4.2.10 Secondary Auxiliary 239 4.5 Universal Affixes 260 4.2.11 Reduplication in Verbs 240 4.5.1 -ka 261 4.2.12 Co-occurrence Restrictions between 4.5.2 -wa 262 any verb root and the Auxiliaries 243 4.3 Particles 4.5.3 -ye 263 243 4.3. I y! 244 4.5.4 -te/-.Je 263 4.3.2 a 245 4.6 Ordering of Affixes in Nominala 265 4.3.3 kama(yy.) - 245 Syntax 270 4.3.4 maqa 247 5.1 Discussion of "Sentence" 271 4.3.5 manal)g8 247 5.2 The Simple Sentence 273 • 1 ~~'. . 5.3 Non-verbal Sentences 274 '...1.. ' 5.4 Word Order 276 .\.•:I . . - " •f,;-' - i 5.S Admissible deletion and possible ambiguity 277 1.1 Linguistic Type. f' 5.6 Possession 281 L 1.1 Muwinypata is an Australian Aboriginal language e .. ,. l 5.7 Comparison 283 spoken (as a first language) by about five hundred people t;~ 5.8 Deri.vatiou 283 mostly at Port Keats. N.T. The language makes up a separate i' 5.9 Relative Clauses 287 language family in itself, apparently unrelated to its .~ 5.10 Other Subordinate Clauses 292 linguistic neighbours [Wurm 1972:122]. It is not a typical ,,.~~. 6 Semantics 294 Australian language although it is typical for its area. 'I 6.1 Noun-classification 294 ... ';';, \;. 6.2 Membership of Noun Classes 305 The languages of the north-west of Australia are f -. '~~" 6.3 Loan Concepts 312 typically prefixing, and, either dual classifying or ",JI; 6.4 Metaphorical Naming using Body Part Terms 314 multiple-classifying [Capell 1940:244-5]. The Daly Family · *,. ~ 6.5 Verb Semantics 316 languages to the north and east of MuJinypata all show noun ~.;-: Appendix 1 Sonagrams 320 classification of some sort, generally multiple classificati~ Appendix. 2 Auxili.ary Paradigms 325 [Tryon 1974:289~ 293-4]. To the south of the traditional Intransitive-type 326 territory of the MUJinypata is the Djamindjungan Family; this Transitive-type 363 has four members one of which, Nungali~ is said to be multiple- Appendix 3 Reflexive and Reciprocal Paradigms 384 classifying while the other three are non-classifying [Bolt, Appendix 4 Texts 394 Cleverly, Hoddinott 1970:593; Worm 1972:122]. Languages of References 435 the north-west are characterized by verb morphologies with a large number of form classes and by the use of auxiliaries , [Birk 1974]. [CapeU and Elkin 1937], [Coate and Oates 1970], ( [Love 1938], [Metcalfe 1972], [Tryon 1974] et a1. MUJ!nypata is a multiple-classifying, prefixing language .~ with complex verb morphology including auxiliaries. It is thus typical for its area - the more so since the writer's own fieldwork ~ndicates that Djamindjung, a member of the Djamindjungan family bordering on Mu~inypata, is multiple classifying despite the statement to the contrary above. 2 3 1. 1.2 Phonologically MUJinypata is quite unusual in singular. dual. paucal and p~ural in all perSons. First Australia in having a voiced/voiceless distinction in stops. person has an inclusive/exclusive distinction in dual and Essentially [2.1] there is a fiv~-place stop-nasal series: paucal numb~rs. There is a two way distinction in duals bilabial, apico-a~veolar. apico-domal, lamina-palatal and and paucals between groups made up of members of the same dorsa-velar. There are three laterals corresponding to subsection and groups not so composed. There is a further the non-peripheral places of articulation of the stop- distinction for the latter between groups which are all nasal series. In these two respects at least, MuJinypata masculine and those which are not. Third singular pronouns follows Australian phonological patterns [Dixon 1970]. show a distinction between masculine and feminine. All There are two semiVowels, labia-velar and palatal and pronouns may inflect in an absolutive-ergative pattern as three rhotics: a resonant and trill with apica-alveolar do nouns. articulation and an apico-domal resonant. There is a four-vowel system, the two parameters being relative 1.1.4 Nouns, noun-classifiers. pronounS (and adject.ives) highness veraus-1owness and relative frontness versus may appear with case-suffixes for absolutive, ergative, backness. In initial position the language allows any instrumental and dative case functions. Other functions of the stops and nasals, the semivowels. the apico-alveolar are signalled by prepOSitions, post positions, or both lateral, l~ and very occasionally the apico-alveolar resonant together~ or, by word order. rho tic , .J. Three words begin with a vowel viz. awu "no"; a "or", possibly a loan word from EngliSh; eynY dYal 1.1.5 Verbs have an extraordinarily rich morphology. "angel", certainly an English loan word. Words may end in Principal syntactic functions are Signalled by cross most consonants or in a two-member cluster. Intervocalically referenCing bound pronouns. The~e pronouns appear in a two and thre~member consonant clusters occur. The assignment nominative-accusative pattern. Indirect objects are also of stress is r~ther complicated and will be treated later [2.2]. cross-referenced. There are a large number of verb classes Words may have anything from one to fifteen (or more) syllables. conjugated by auxiliary somewhat like Ngarinyin [Coate and Oates. 1970:54J but more similar to Malak Malak [Birk 1974] 1.1.3 The parts of speech (word classes [3·] ) are as in tha t any verb may occur with a number of auxiliaries follows: noun, noun-classifier, prono':m. adjective, verb, (provided it ,is semantically feasible). adverb, adjective/adverb qualifier. particle. and interjection. There are ten noun-classifiers [3] which precede the noun 1.1.6 Sentence modification 1s carried out by particles they classify. The pronouns distinguish four numbers: which convey temporal, aspectual and locational information. , 5 1.2 Dialect Situation. 1.2.1 The term. IIMuJinypata LanguageIT is. a cover term for Furthermore the use of the term three separate speech forms. The usage follows that of 4a 4iminin the local popu1ation who say IThere are three "ways of place gravel -talking": MUJinypata, MUJinykuJ8 and Ml.i..finygi.minin but gravel country" these are all MU..linypata I. The speakers of the Mu.Jinypata It dialect predictably claim that their dialect is the best. was not considered appropriate for the traditional dialect clearest and most euphonious of the three. The name area of MU.Jiny9iminin. itself is suggestive of this bias: 1.2.3 Informants offered some meta1inguistic suggestions mU.JinY-pata ~ut these have not been closely checked and cannot be language-good held as being reliable. MU.Jinygiminin was said to be so "g09d language" named because it was a "rough" language: The other two dialects come out as: pewetipeweti ~inipunY giminin mUJinY- kuJa rough like gravel "rough like gravel" language-water !'water language" MU.JinykuJa. however, was so-called because it is "softU and (see also Falkenberg [1962:13] ). mU-linY- 9-iminin language-gravel The other metalinguistic parameter employed was "heavy" '-' "gravel language" vs. "light!!. "Lightnessll was considered a virtue in a language; MUJinypata is 1l1ightll• MU.Jinyku.Ja and MUJiny4iminin 1.2.2 It is difficult tc? say why .th ese descriptions have were said to be "heavy" and thus not so easily understood • , been used. These names may suggest the geographical locatio~ Informants complained of the writer's lack of clarity in of the tradit~onal areas in which the dialects were spoken: diction for the reason that his speech (often descrihed as MU.JinykuJa has a border on the water while MU.Jiny4iminin is Broad Australian) was "heavy - like a Yankee"!. inland (although the country is not notable for its gravel). 6 1.2.4 The d~str1butian of dialects is given in the map (1.2.4.A). For the distribution of clans and hordes see Falkenberg [1962:pp21ff. esp. 23]. 1.2.5 What little has been collected of MuwinykuJa and MUJiny~iminin indicates that they are very closely related (over 80% shared vocabulary, there being similar percentages between MUJlnypata and the other dialects), It has not been Port Keats possible to elici~ reliable names for the three dialects in MUJiny4iminin the MUJinykuJa or MUJiny~iminin. The practice through this work will be to use the name of a language in that language. Nangu MUJinypata - Throughout the description (unless otherwise stated) the term MU.Jiny-pata will apply to the HUJinypata dialect of the Hu.linypata language. (See also [1.4] ). 1.2.6 Presently most speakers of MUJinypata live at Port Keats Mission (formerly MUJinY91minin territory). There are some speakers at Kununurra W.A. and a few speakers working at stations to the south. MU.linypata acts as a lingua N t franca for the Port Keats area: nearly the total population of Port Keats (c.IOOO) understand and regularly use MUJlnypata even if it is not their first language. Perhaps a third to a half of the population have MUJinypata as their "first language" although the term is dubiously applied when 1.2.4.A Language Map of Port Keats.1}\rea. there is much tribal intermarriage and very many people are bi- or multi-lingual. 7 e 1.3 Surrounding Languages. 1.3.A Percent~ges of shared cognates between MUJinypata 1.3.1 No detailed discussion of the surro90ding languages and its linguistic neighbours: Djamindjung and the Daly and their relationship ~ith Mu~inypata will be given here. Family. MU.Jinypata Tryon [1974] gives an account of the Daly Family Djamindjung 9.0 languages which covers the northern and eastern linguistic Mullukmulluk 3.5 neighbours of MUJinypata. Tyeraity 5.8 Matngala 3.5 Yunggor 1.8' The southern linguistic neighbour, the Djamindjungan Kamor 4.1 _Family. is described by Cleverly [1968]; Bal~. Cleverly Marithiel 8.7 and Haddinott [1970] and Bolt, Hoddinott and Kofod [1971]. :~I Marityaben 10.5 Hare Ammu 9.9 Haridan 8.2 1.3.2 Lexically MUJinypata is not closely related to its Maramanandj i e.7 neighbours. Tryon (personal communication) indicates Harengar 10.5 :'rI Maranunggu 8.1 that MUJinypata shows few cognates with any of the Daly Ami 8.1 languages. Street [1973b] conducted a survey at Port Keats Handa 7.6 in which the~rinken and Tyemeri subgroups of the Daly Family Pungupungu 5.2 Wadyiginy 5.2 [Tryon: 1974, xiii] and Dj~mindjung are compared with Batyamal 5.8 ;~1 Mu~inypata. The writer has compared MUJinypata vocabulary Ngangikurrunggurr 9.9 from his own field notes with TryonTs word lists for the ~t Ngengotneri 9.9 Daly Famdly languages [1974:267-286] and with vocabulary from Cleverly's [1968] Djamindjung vocabulary. This is Percentages of. shared cognates within the Daly Family given in the table: 1.3.A. The results essentially agree is given in Tryon [1974:xiv]. Predictably the percentages with those of Street. are marginally higher for those languages which are .... geographically closer .to MUJinypata • :