VERBAL MORPHOLOGY OF THE MAIN SENTENCE IN NINAM OF UPPER MUCAJAÍ, A YANOMAMI LANGUAGE by VICTORIA ELIZABETH INFANTE PEÑA Bachelor of Education in Language and Literature, Universidad Nacional Experimental de Guayana, 2020 Thesis Submitted in Partial Fulfillment of the Requirements for the Degree of MASTER OF ARTS IN LINGUISTICS in the FACULTY OF GRADUATE STUDIES TRINITY WESTERN UNIVERSITY © Victoria Elizabeth Infante Peña, 2024 APRIL, 2024 i Acknowledgements A research project is the result of personal inspiration and effort combined with the support offered by the people who esteem us, without which we wouldn't have the strength that encourages us to grow both professionally and personally. That's why I consider it necessary to recognize and thank, first and foremost, Teus (God) because all wisdom and every perfect gift comes from Him, and secondly, all those who accompanied me during the development of this project: To Pedro and Olga, for fostering the love for studying since Luisda, Migue, and I were children. To Raquel and Luisa, for browsing libraries in Caracas and La Guaira. To Sean Allison for his support and for the encouragement he gave me to continue studying since the morphosytax course. Definitely, Morphosyntax is morphosinating. To Steve and Thiago for their valuable comments and corrections for the final submission. To Xakalina, Professora Hosa, and Kilo (lina) for their help during the data collection and analysis. To MEVA for the shared data. To ALEM, TSCO, and CanIL for their contribution to my academic formation. To the Ninam for sharing their language. And, of course, to Juan, for his patience, for listening to me every time I wanted to verify if what I was saying made sense, and for telling me "you can do it.” Thank you all. ii Abstract The Yanomami are a linguistic family located in the Orinoco-Amazon jungle, between Venezuela and Brazil, in South America. The family is characterized by a highly polysynthetic typology, especially related to the verb. This work describes the morphemes that comprise the verbal complex in the main sentences of Ninam, a Yanomami language spoken in the Upper Mucajaí. The work begins with a brief historical review related to the Yanomami people with special emphasis on the Ninam of the Upper Mucajaí. A summary of the important aspects of phonology is given, along with the orthography used for the presentation of the data. The presentation of the morphological features of the verbal complex begins with a look at the types of verbs, followed by a description of the preverbal morphemes, which mark subject, object, simultaneity and evidentiality. The postverbal morphology marks valency changes, adverbials, aspect, directionals, locatives, and tense. iii Contents Acknowledgements .................................................................................................................. i Abstract.................................................................................................................................... ii List of tables..............................................................................................................................v List of figures .......................................................................................................................... vi List of abbreviations ............................................................................................................. vii List of abbreviated references............................................................................................... ix 1. The Yanomami linguistic family ....................................................................................1 1.1 Location, people, and languages ................................................................................1 1.2 Typological profile.....................................................................................................2 1.3 Ninam and the variety spoken in the Upper Mucajaí River region ...........................4 1.4 Early literature ...........................................................................................................7 1.5 Corpus used in the analysis ......................................................................................10 1.6 Purpose of this work and overview of the following chapters.................................11 2. Remarks on phonology ..................................................................................................12 2.1 Consonants ...............................................................................................................12 2.2 Vowels .....................................................................................................................16 2.3 Syllable structure .....................................................................................................18 2.4 Word and stress ........................................................................................................18 2.5 Orthography .............................................................................................................19 3. The verb stem in Ninam of Upper Mucajaí .................................................................21 3.1 Stative verbs .............................................................................................................21 3.2 Dynamic ...................................................................................................................23 4. Preverbal morphology ...................................................................................................26 4.1 Subject and object marking ......................................................................................26 4.2 Simultaneous ............................................................................................................35 4.3 Evidentiality .............................................................................................................36 5. Postverbal morphology..................................................................................................41 5.1 Pluractional =a~=la ................................................................................................41 5.2 Valency change ........................................................................................................43 5.3 Adverbials ................................................................................................................47 5.4 Aspect marking ........................................................................................................48 5.5 Middle voice =o.......................................................................................................56 5.6 Verbal directionals and locatives .............................................................................57 5.7 Tense, aspect, mood, infinitive and comitative markings ........................................63 5.8 Pronominal reference to the ergative 3rd dual and plural .........................................69 Conclusions .............................................................................................................................72 References ...............................................................................................................................74 Appendices ..............................................................................................................................77 5.9 Appendix 1: The story of the spider monkey (mythical narrative) ..........................77 iv 5.10 5.11 Appendix 2: The word (narrative) ...........................................................................79 Appendix 3: Cassava drink (procedural) .................................................................82 v List of tables Table 1 – Ergative-absolutive alignment among Yanomaman languages (adapted from Ferreira, Machado & Benfica 2019: 120) ....................................................................3 Table 2 – Classifiers (adapted from Ferreira, Machado & Benfica 2019: 124) ..................3 Table 3 – Yanomami villages and population of Upper Mucajaí (Missão Evangélica Caiuá 2023: 7). .......................................................................................................................7 Table 4 – Consonants in Ninam of Upper Mucajaí (Adapted from Borgman 1959a: 2) ..12 Table 5 – Contrast between phonetically similar consonants ...........................................13 Table 6 – Phonemic consonants of Ninam of Upper Mucajaí ..........................................16 Table 7 – Phonemic vowels of Ninam of Upper Mucajaí (Borgman 1959a: 2) ...............16 Table 8 – Contrast between oral vowels ...........................................................................17 Table 9 – Contrast between oral and nasal vowels ...........................................................17 Table 10 – Phonemic vowels of Ninam of Upper Mucajaí ...............................................18 Table 11 – Syllable patterns found in Ninam of Upper Mucajaí (Adapted from Borgman 1959a: 12) ..................................................................................................................18 Table 12 – Syllable patterns in Ninam of Upper Mucajaí.................................................18 Table 13 – Current orthography system of Ninam of Upper Mucajaí ..............................20 Table 14 – Position for the proclitics in the verbal complex of Ninam of Upper Mucajaí26 Table 15 – Pronominal markers and Pronouns in Ninam of Upper Mucajaí ....................27 Table 16 – Sequence for subject and object marking in Subsystem A .............................28 Table 17 – Sequence for subject and object marking in Subsystem B. ............................30 Table 18 – Sequence for subject and object marking in Subsystem C. ............................31 Table 19 – Position for the enclitics in the verbal complex of Ninam of Upper Mucajaí 41 Table 20 – Morpheme breaks in the locatives of Ninam of Upper Mucajaí .....................59 Table 21 – Locatives of the horizontal plane in Ninam of Upper Mucajaí .......................59 Table 22 – Locatives of the vertical plane in Ninam of Upper Mucajaí ...........................61 Table 23 – Past tense continuum in Ninam of Upper Mucajaí .........................................63 vi List of figures Figure 1 – Map showing the dialect division ................................................................... 2 Figure 2– Map showing Ninam’s dialects according to Migliazza (1972) ...................... 4 Figure 3 – Current distribution of the Yanomami languages in Brazil and Venezuela. .. 5 Figure 4 – Spectogram and intonation of the phonological word kayehe...................... 27 Figure 5 – Spectogram and intonation of the phonological word yehekẽawei .............. 28 Figure 6 – Spectogram and intonation of the phonological word wiitihi ...................... 33 Figure 7 – Spectogram and intonation of the phonological word tihipolalei ................ 34 Figure 8 – Map showing the villages around the Upper Mucajaí village ...................... 60 vii List of abbreviations ?? 1 2 3 ADV AN ANA ASP CAUS CLN CNJ COMIT CONCS CONS COP DEM DIR DL DYN ERG EVID EXPLV FUT HAB HEAR HOD IMM IMP IN INCEP INSTR ITIV LEV LOC NEG NMLZ OBJ PLANE PL PLU PFV POSS POST Category under study First person Second person Third person Adverbial notion Animate Anaphoric pronoun Aspect Causative Nominal Classifier Conjuntion Comitative Concessive conjunction Conservative Copula Demonstrative Directional Dual Dynamic Ergative Evidential Explanative conjunction Future Habitual Hearsay Hodiernal past Immediate Imperative Inanimate Inceptive aspect Instrumental Itive Level axis for locatives Locative Negation Nominalizer Object Plane axis for locatives Plural Pluractional Perfective Possessive Positional viii PRE.HOD PRS REC. PST1 REC.PST2 RECIP REFL REM.PST RES SEQ SG SIM STV SUBJ UNSP VEN VIS Prehodiernal Present Recent past tense 1 Recent past tense 2 Reciprocal Reflexive Remote past Resultative aspect Sequential Singular Simultaneous Stative Subject Unspecified Venitive Visual ix List of abbreviated references SA65 CS71 CS02 CS02b CS04 Albright (1965) Swain (1971) Swain (2002) Swain (2002b) Swain (2004) 1 1. The Yanomami linguistic family 1.1 Location, people, and languages Yanomami – also called Yanoama, Waicá, Xirianá, Yanomama, and Guarahibo – is a linguistic family located in the Orinoco-Amazon rainforest between Venezuela and Brazil in South America. The area comprises national parks on the Venezuelan side and the Yanomami Land on the Brazilian side, covering roughly 192,000 km2 (74,131.61 sq.mi.). As some Yanomami communities are difficult to access, it is challenging to measure their population accurately. According to the Instituto Nacional de Estadística (2011), the last official census conducted in Venezuela failed to provide an accurate estimate of the Yanomami population in the country, particularly concerning the number of villages that are known to exist. At that time, 11,431 Yanomami were counted. This number corresponds to more accessible villages and includes the Sanema and Ninam groups, which are classified as part of the Yanomami family but listed separately. On the Brazilian side, the number is 26,780 Yanomami (Povos Indígenas no Brasil 2023). During the 20th century, it was a matter of debate among linguists as to whether Yanomami varieties formed a continuum of dialects or whether they were, in fact, different languages. Borgman et al. (1965), as well as Aikhenvald & Dixon (1999: 341–342), considered that it was indeed a dialect continuum. Furthermore, Aikhenvald & Dixon’s study indicated that the dialects were lexically proximate, having “the average number of lexical cognates on a 200-word list in 70 – 80%” and also sharing 95% of their grammatical morphemes (Aikhenvald & Dixon 1999: 345). Figure 1 shows the division of the dialects suggested by Borgman et al. (1965: 2), note that the term Waica is one of the names used in early studies. On the other hand, Migliazza (1972: 34), Migliazza (1985: 27), and Campbell (1997: 205) suggest that the Yanomami family is composed of four different languages: Yanam, Sanuma, Yanomámɨ, and Yanomamö. This proposal prevailed for almost forty years. Later, Ferreira (2011) stated the existence of five languages: Ninam (Yanam), Sanöma (Sanuma), Yanomamɨ (Yanomámɨ), Yanomam (Yanomamö), and Ỹaroame. This five-language division resembles the classification proposed by Ramirez (1994). Even though recent studies from Ferreira, Machado & Benfica (2019) add Yãnoma to make a total of six languages, this thesis 2 will only consider the classification of five languages since this is the currently accepted categorization by the scientific community, such as Ethnologue and Glottolog. Figure 1 – Map showing the dialect division proposed by Borgman et al. (1965: 2) 1.2 Typological profile In general, Yanomami languages have about 7 vowels, each of which has a nasal counterpart, and 12 consonants (Campbell 1997; Aikhenvald & Dixon 1999; Ferreira 2017; Ferreira, Machado & Benfica 2019; Eberhard, Simons, & Fennig 2023). The syllabic structure commonly used is a consonant followed by a vowel (CV), although some languages of the family allow the formation of CCV syllables. The stress is predictable and always falls on the penultimate syllable, except for words with CVC syllables in the final position, in which case, the stress falls on the last syllable. 3 Another important characteristic of Yanomami languages is their morphosyntax. In this regard, these languages are known for being highly polysynthetic languages and for having a preferred subject-object-verb order (Aikhenvald & Dixon 1999: 346, 350). The two major grammatical categories are nouns and verbs, which host a large number of morphemes containing grammatical and semantic information in a single phonological word. In terms of the alignment system, Derbyshire (1987) states that the ergativeabsolutive system is very common in the Amazonian area. Aikhenvald & Dixon (1999) and Ferreira et al. (2019) confirm that all the Yanomaman languages present this alignment system. Table 1 indicates that, among the Yanomami languages, the ergative marker is a cognate form of -nɨ. Table 1 – Ergative-absolutive alignment among Yanomaman languages (adapted from Ferreira, Machado & Benfica 2019: 120) Sanöma Ninam Ỹaroamë Yanomamɨ Yanomam Ergative =nɨ =n ~ =nɨ =nə [nɨ] =nɨ =nə ~ =nɨ Absolutive Ø Ø Ø Ø Ø The use of nominal classifiers is also an interesting feature of these languages. Borgman (1990), Derbyshire & Payne (1990), and Aikhenvald & Dixon (1999) propose that nouns in the Yanomaman languages are divided into classified and non-classified nouns. The main difference between the two types of nouns is related to their morphological complexity. The non-classified nouns are morphologically simple; this means that this type of noun is formed by a free word. Meanwhile, classified nouns are morphologically complex, formed by a root and a noun classifier. Table 2 shows examples of nominal classifiers in the Yanomami languages. Table 2 – Classifiers (adapted from Ferreira, Machado & Benfica 2019: 124) Classifier Gloss =(t)u/=uku/=up(ə) liquid Example water banana porridge Ninam Ỹaroamë ma =up maa u katha =uku koata uku Yanomamɨ mau u kurata ukɨ =ma stone, hard material maa kë ma tree maa =ma mai =ma lolo na=hi oroxi =hi wi =tihi maa ma =hi/=ti stone açai cacao tree cashew tree tree (gen) poro na=hi purumë =hi hii të=hi pohoroa =hi wito =hi hii =hi 4 1.3 Ninam and the variety spoken in the Upper Mucajaí River region According to Ferreira et al. (2019) , the Ninam – also known in the literature as Yanam, Xirixana, Xiriana, Waika, Xiliana, Shiliana, Xilixana or Shilishana – migrated from the Sierra Parima mountain range in a northerly direction. However, various conflicts with other groups, such as the Ye’kwana and Maku, motivated the Ninam to move southward. This migration gave rise to different variations of the Ninam language. In his survey, Migliazza (1985: 30–32) categorized the Ninam language into three distinct dialect groups. The northern Ninam inhabited the vicinity of the Paragua region in Venezuela as well as the Uraricaá River in Brazil. The central Ninam resided near the middle region of the Mucajaí River in Brazil, and the southern Ninam were situated along the tributaries of the Apiaú and Ajarani rivers. Figure 2 shows these three groups in the abovementioned locations. Figure 2– Map showing Ninam’s dialects according to Migliazza (1972) (adapted from Early & Peters 2000: 10) Due to studies in phonology and syntax carried by Albright (1965) and Migliazza (1972), Migliazza (1985: 32) explains that Southern Ninam, which was located in the vicinity of the Apiaú and Ajarani rivers – c.f. Figure 2–, is different from Central Ninam and Northern Ninam. Therefore, it may not be a dialect of the Ninam language. However, at that time, no specific studies had been conducted on this particular variant. 5 Figure 3 – Current distribution of the Yanomami languages in Brazil and Venezuela (Tinoco 2023). Later, Ramirez (1994: 24–25) acknowledged Southern Ninam as a distinct group by labeling it as Language A – or Language of the Ajarani River. Nevertheless, this proposal did not gain widespread acceptance due to the limited linguistic information available at the time. 6 It was not until 2011, when Ferreira presented linguistic data and a grammatical outline, that Yãroame was recognized as a distinct language rather than a dialect of Ninam. This linguistic distinction is depicted in Figure 3, which illustrates the current distribution of the Ninam language and the other four Yanomami languages in both Brazil and Venezuela. The Ninam who settled near the Mucajaí River remained in relative isolation until 1956 when they decided to contact other groups in the search of metal goods. The first expedition was unsuccessful, but in 1957 “a group of men descended the Mucajaí in canoes and contacted some Brazilians. The Ninam traded arrows and canoe paddles for a few axes and knives and then returned upriver. Shortly afterward another group of twenty men and two women made another trip” (Early & Peters 2000: 32). Further on, in 1958, missionaries from the Unevangelized Field Mission (UFM), currently known as Missão Evangélica da Amazônia (MEVA), encountered them (Migliazza 1985; Early & Peters 2000). In 1990, a cluster of 40 individuals who belonged to the Ninam of the Upper Mucajaí River community made the decision to migrate and relocate to the Uraricoera River. As a result, they established what is presently referred to as the Central Ninam, according to Ferreira, Machado & Benfica’s (2019) categorization. The migration led to the emergence of a new group, which left the original Ninam population residing in the Mucajaí region, currently referred to as the Southern Ninam. However, it should be emphasized that there is no linguistic information available that proves that Central Ninam is a dialect different from the one spoken in the Upper Mucajaí region. In fact, the Southern Ninam speakers refer to Central Ninam only as “a variety spoken more rapidly than Southern Ninam” (Ferreira et al. 2019: 42). In the span of the past five decades, the Ninam located in the Upper Mucajaí River has had intense contact with Brazilians. Ferreira et al. (2019) describe this contact as the result of the establishment of educational institutions, the arrival of missionaries, the construction of airstrips, and the provision of medical facilities within the villages. In addition, due to the distance from small towns in Brazil, the Ninam of Upper Mucajaí worked in farms and developed a sponsorship arrangement with Brazilians. This sponsorship involved sending young Ninam children to be raised by Brazilian families, thereby further strengthening the interconnections between the Brazilian population and the Ninam community of the Mucajaí region. 7 According to Ferreira et al. (2019), the Ninam of Upper Mucajaí total approximately 900 speakers. However, more recent information indicates a lower number. The Relatório Missão Yanomami (2023) indicates 685 people, and a slightly lower number – 655 people – is reported by the Missão Evangélica Caiuá (2023), which makes this language one of the smallest groups of the Yanomami family. Although the latter indicates a smaller number of Yanomami in the Upper Mucajaí region, this report provides information on the names of the villages and the number of inhabitants concentrated in the Upper Mucajaí, c.f. Table 3. Table 3 – Yanomami villages and population of Upper Mucajaí (Missão Evangélica Caiuá 2023: 7). Village Population Âpolihipii 37 Caju 17 Caju 2 44 Ilihimakok 36 Kuisi (Buritis - Kasilapai) 23 Lasasi (Ex Mário Missão) 140 Porapi 22 Pewaú 76 Thoribi 48 Uxiu 123 Waikaú 32 Milikowaxi Do Geladeira 34 Pedral 23 Total 655 To ensure clarity and to account for historical naming conventions, the groups mentioned hereafter will be referred to using specific geographical references rather than cardinal points. Therefore, the variant of interest in this research will continue to be identified as Ninam of Upper Mucajaí providing a more precise and contextually relevant designation. From §3 onwards, the name Ninam will also be used to refer to the Ninam language spoken in the Upper Mucajaí region. 1.4 Early literature The Yanomami group has garnered significant attention within academic circles as one of the most researched communities in the Amazon area. Conversely, there exists a notable disparity in the extent of exploration dedicated to the Yanomami languages. Unfortunately, this particular domain remains relatively limited, with only a few linguists having ventured into descriptive investigations within this realm. In §1.4.1. I exclusively list works that focus 8 on the grammatical description of Yanomami languages, disregarding other types of publications that may exist. In §1.4.2. I list chronologically all the works related to the Ninam language. 1.4.1 Research on the Yanomami languages The initial documentation of Yanomami languages can be traced back to 1913 when KochGrunberg documented his exploratory journeys in the regions of Roraima and Orinoco. Koch-Grunberg (1923) categorizes the Schirianá language as an isolate and supplements his analysis with a lexicon encompassing approximately 300 words (Koch-Grunberg 1923: 302– 308). A significant milestone in the examination of Yanomami languages occurred sixty years later with Migliazza’s contribution in 1972, which can be regarded as a foundational work in this field of study. Migliazza (1972) provides a description of the intelligibility and grammar of the Yanomami languages. Notably, this scholarly work holds substantial importance in the domain of Yanomami linguistics as it represents the initial recognition of the Yanomami community as comprising four distinct languages rather than mere dialectal variations. In the Venezuelan context, Barker (1979) stands out as the pioneer in studying and describing a Yanomami language. He is referred to as "the Panini of the Yanomami language" (Mosonyi 2014). Barker’s groundbreaking work laid the foundation for an understanding of Yanomami grammar. Furthermore, Lizot (1996) also engaged in the description of other Yanomami languages. On the Brazilian side, Borgman (1990) provides an outline of the grammar of the Sanuma language. Ramirez (1994) undertakes a meticulous study of the Yanomamɨ of the Xamatauteri region, including a thorough examination of its phonological, morphological, and syntactic aspects, and Ferreira (2011; 2017) offers grammatical sketches of the Yarõame and Yanomam languages, respectively, shedding light on their grammatical structures and other linguistic properties. 1.4.2 Research on the Ninam language As previously mentioned, Borgman’s research trajectory involved his study of Sanuma, another Yanomami language. However, prior to this, he conducted preliminary investigations 9 on the Ninam language, which were documented in two unpublished reports. These reports, stemming from the initial expeditions in 1958, present preliminary data on the language, covering aspects of both phonology and morphology (Borgman 1959a) and an initial attempt at the morphological structure (Borgman 1959b). For the second report, Borgman indicates that much of the material was transcribed phonetically due to uncertainties in the phonological realm. However, the most significant contribution of these works lies in the initial attempt to delineate the position of the morphemes, providing valuable insights into the morphological structure of the Ninam language. Albright’s (1965) and Tracy’s (1966) work focus primarily on the phonological aspects of Aikamtheli, a Ninam variation spoken in the Upper Parima River region. Albright’s (1965) research encompasses several key areas, including the definition of the phonological foot within the language, the exploration of intonation contours, and morphophonemics. Tracy’s (1966) research went beyond preliminary investigations and ventured into the realm of grammar. Although the emphasis is placed at the sentence level, Tracy’s work does offer relevant insights pertaining to specific morphemes. The works by Albright (1965) and Tracy (1966) served as foundational references for subsequent linguists, including Swain’s (1971) research of Ninam of Upper Mucajaí. In Albright’s (1970) study, the primary focus is an examination of morphological features within Ninam of Upper Mucajaí. Within this article, Albright specifically addresses grammatical categories that are expressed through what she identified as discontinuous morphemes. According to the author, these unique morphemes serve as a means for Ninam speakers to convey kind of knowledge (evidentiality), source, location, and tense. In the subsequent year, Swain, a colleague of Albright, continued to work among the Ninam of Upper Mucajaí and wrote an unpublished grammar sketch of the language. Given the presence of other studies that primarily address phonological aspects, Swain’s (1971) work prioritizes the examination of clausal and sentence levels and discourse rules. Despite the primary focus on clauses, Swain (1971) includes tables in which morphemes belonging to the verb phrase and their glosses are presented. However, due to the emphasis on clause-level analysis, detailed descriptions of each individual morpheme are not provided. The tables serve as a reference for understanding the composition of the verb phrase in Ninam of Upper 10 Mucajaí, offering insight into the structure and organization of the language at the clause level. In Swain (1975), the research delves into the analysis of narrative structure, specifically focusing on the concept of climax within the Ninam discourse. Fifteen years later, new research about Ninam was carried out by Gómez. In Gómez (1990), the primary focus lies in the analysis of Shiriana, a dialect of the Ninam language spoken in the Upper Uraricoera River region. One chapter within the thesis specifically delves into the morphology of Ninam of Ericó, surveying various aspects such as pronominal forms, kinship terminology, possession, and the noun phrase. Additionally, a chapter in the thesis explores the verb and verbal syntax of this Ninam variant, thoroughly examining the characteristics of verb usage in the language. Later, in 2018, Gómez also described nonverbal predications in the same language. 1.5 Corpus used in the analysis The corpus of data used for this study comprises ten distinct texts and elicited sentences, for a total of 616 sentences. Eight out of the ten texts and the sentences were collected through the process of elicitation, conducted over a period spanning from June 2022 to June 2023. The data collection involved three male Ninam native speakers, aged between thirty and forty. Before engaging in the sessions, written consent was obtained from each participant. During the elicitation sessions, participants were prompted to share narratives and procedural texts, which were then recorded for analysis. Subsequently, the recorded texts were transcribed. In addition to the elicited data, two additional texts, one narrative and one procedural, a list of 1,670 sentences, and a list of 1,495 words were obtained from data collected and transcribed by Missão Evangélica da Amazõnia (MEVA). By incorporating these various sources of data, encompassing both elicited data and those obtained from the MEVA database, this study has aimed to construct a comprehensive and representative database to carry out an in-depth analysis and examination of the linguistic properties of the verbal complex of Ninam of Upper Mucajaí. Data obtained from MEVA documents used in this work will be referenced by using the abbreviations listed in the List of abbreviated references, page x of this work. 11 1.6 Purpose of this work and overview of the following chapters Based on the literature available, the Ninam language spoken in the Upper Mucajaí region has received considerable attention at various linguistic levels, including phonology, clausal structure, sentence formation, and discourse analysis. However, a significant gap remains in terms of a detailed morphological description of this language. While previous studies have shed light on several linguistic aspects, a comprehensive understanding of its morphology is still lacking. Thus, the primary objective is to address this gap by providing an initial description of the verbal morphology of Ninam of Upper Mucajaí, with a specific focus on the morphology of the verbal complex in the main sentence. By following the principle of describing the language in its own terms and by understanding that languages have their own structure (Haspelmath 2008: 1), I aim to construct the set of concepts needed for the description of Ninam. Therefore, this description is framed in the framework-free approach proposed by Haspelmath (2008). To achieve this goal, the sections of this paper are organized as follows: Firstly, an overview of the language’s phonology will be presented, drawing upon existing research and findings. The subsequent parts of this study will delve into the analysis of verbal morphology. Through this comprehensive analysis, the aim is to contribute to a deeper understanding of the intricate morphological features inherent in Ninam of Upper Mucajaí. 12 2. Remarks on phonology This chapter focuses on the phonological characteristics of the Ninam language in the Upper Mucajaí region. The primary source for this examination is the research conducted by Borgman (1959a), which serves as the main resource available on this specific variant of Ninam. Although this chapter does not intend to present an exhaustive phonological analysis, the information provided by Borgman (1959a) will be compared with the data collected for this study, as well as with other literature related to the Yanomami language family. 2.1 Consonants Borgman’s (1959a) study constitutes a preliminary investigation based on data collected during early contact with the Ninam community of the Upper Mucajaí. As mentioned in §1, Borgman subsequently focused his attention on another Yanomami group, which limited the depth of the phonological analysis of the Ninam variant of the Upper Mucajaí compared to the descriptions offered by Albright (1965) and Tracy (1966). Borgman’s (1959a) phonemic inventory presented in Table 4 shows the phonemes postulated from the data collected between 1958 and 1959. This inventory identifies a total of thirteen consonants. The number of consonants corresponds to the expected number of consonant phonemes found in the literature related to other Yanomami languages (Aikhenvald & Dixon 1999; Campbell 1997; Ferreira et al.2019; Eberhard et al. 2023). Table 4 – Consonants in Ninam of Upper Mucajaí (adapted from Borgman 1959a: 2) Bilabial Interdental Alveolar Palatal Velar Plosive Voiced b d Voiceless k h Aspirated t kh Affricate Voiced [d͡ʒ] Fricative Voiceless θ ʃ Nasal Voiced m n [ɲ] Liquids [l] Voiced ɾ Glide Voiced w j Glottal h 13 In comparing the data presented by Borgman (1959a) with more recent data, some differences have been noted regarding the natural classes. In relation to the obstruents, Borgman’s (1959a) study indicates that the interdental fricative fluctuated with an alveolar fricative, suggesting free variation. However, data collected for this study did not show the presence of the interdental fricative; instead, only the alveolar fricative was found in the newer data. It was also noted that the aspirated velar plosive is missing from the newer data. Table 5 shows the contrast for the obstruents, nasals, and labiovelar glide. Table 5 – Contrast between phonetically similar consonants b/m [baa] ~ [paa] ‘hoplias malabaricus’ [boko]~ [poko] [maa] ‘rain’ [moko] b/w [waka] [baka] ~ [paka] ‘priodontes maximus’ ‘hole’ b/d [hubu] ~ [hupu] [hudu] ~ [hutu] ‘hold’ ‘field’ d/n [naʃo] [daaʃo] ~ [taaʃo] d/tʰ ‘branch’ ‘girl’ [paʃo] ~ [baʃo] [waʃo] ‘robust capuchin monkey’ ‘let’s eat!’ ‘cracked’ ‘let’s see’ [nokoa] [doko] ~ [toko] ‘lick’ ‘run away’ [hudu] ~ [hutu] ‘field’ ‘circle’ [hutʰu] ‘sap’ [hoɾodo] ~ [hoɾoto] [hoɾotʰo] d/s [kadi] ~ [kati] [kasi] ‘in line’ ‘lays on its side’ [dika] ~ [tika] [sikãmõɾĩ] ‘sting’ proper name d/d͡ʒ [dubu] ~ [tupu] [d͡ʒubu] ~ [d͡ʒupu] ‘shake’ ‘carry with the fingers’ [diʃa] ~ [tiʃa] [d͡ʒiʃa] ‘peel off’ ‘pound with a pestle’ k/w [ka] [wa] ‘already’ 2SG [kaka] [waka] ‘split from end to end’ ‘priodontes maximus’ tʰ/h [tʰai] [hai] ‘small’ ‘left’ [tʰa] [haa] ‘to do’ ‘to go’ h/ʃ [hipo] [ʃid͡ʒo] ‘to have’ ‘to fight’ ʃ/s [kaʃi] [kasi] ‘to put in and take out’ ‘lays on its side’ [ʃima~ʃimo] [simo] ‘to want’ ‘starch’ m/n [mãɾĩ] ‘dream’ [ʃima~ʃimo] [nãɾĩ] ‘Bixa orellana’ [ʃina] ‘to want’ ‘tail’ ‘bowl’ 14 2.1.1 Comments on the alveopalatal affricate and the palatal nasal According to Ferreira et al. (2019), in Proto-Yanomami, the palatal glide had a palatal nasal allophone preceding a nasal vowel. This process can be explained with the rule in (1). (1) /j/ à [ɲ] / __ Ṽ à [j] / elsewhere This distribution is still maintained in other varieties of Yanomami. However, in Ninam of Upper Mucajaí, a process of strengthening occurred, resulting in the alveopalatal affricate [d͡ʒ], instead of the palatal glide. According to Borgman’s (1959a) data, both sounds coexisted at that time. However, with the new data, only the palatal affricate and the palatal nasal occur in the syllable onset position, in oral and nasal contexts, respectively. The palatal glide remains as a remnant, appearing in syllable codas. Based on this information, it is possible to postulate a continuation or expansion of the original rule, presented in (2). (2) /j/ à [d͡ʒ] / $__ V à [ɲ] / $__ Ṽ à [j] / __ $ Another approach was adopted by Ferreira et al. (2019) in their analysis, where they argue that the alveopalatal affricate and the palatal nasal should be considered phonemic in the language. According to Ferreira et al. (2019), the Ninam and Ỹaroamë languages developed a change in which the sound [j] was turned into [d͡ʒ] in oral contexts, losing its similarity to [ɲ], and both sounds became phonemic in these languages. Although there are still words with the sounds [d͡ʒ] and [ɲ] in the vocabulary of these languages, this distinction is no longer evident in morphophonological processes. This suggests that the original phonological process that created this distribution is no longer active in Ninam and Ỹaroamë and is considered dormant. However, given the clear distributional patterns observed, I am hesitant to endorse this view. So far, no data has been found where the affricate occurs in contexts with nasal vowels, nor has the palatal nasal been found in purely oral environments. While the possibility of considering the alveopalatal affricate and the palatal nasal as phonemes is not entirely dismissed, further investigation is needed to explore alternative 15 explanations and examine additional instances to shed light on the precise distribution and phonemic status of these sounds. 2.1.2 Comments on the liquids In relation to the liquids, it has been established that [l] and [ɾ] occur in free variation, as in all other Yanomami languages (Ferreira et al. 2019: 116). However, the data from Borgman (1959a: 10) suggest a possible complementary distribution. Borgman describes the [l] as an allophone of the alveolar flap [ɾ]. He also states that it occurs word-initially, between [a] and [o], or in a consonant cluster. From the newer data, it was observed that the lateral liquid only occurred word-initially, while the flap predominantly occurred in the medial position and in consonant clusters. Example (3) shows the contrast between an alveolar plosive and the flap, as the lateral will not occur in this position. (3) [hudu] ~ [hutu] [huɾu] ‘field’ ‘otter’ For Ninam of Upper Mucajaí both sounds can occur word-initially as in [laɾe] ‘put on’ and [ɾoɾo] ‘Pteronura brasiliensis.’ However, only the alveolar flap occurs in the wordmedial position as in [huɾu] ‘otter.’ Therefore, the alveolar lateral and alveolar flap will be considered phonemes, as postulated by Ferreira et al. (2019); however, I would not postulate free variation but a neutralization relation in the medial position. This is because there is no evidence in my data that the lateral alveolar occurs in the medial position. This position is exclusive to the alveolar flap. Thus, the phonemic chart that best represents the consonants in the data collected for this study is the one presented in Table 6. 16 Table 6 – Phonemic consonants of Ninam of Upper Mucajaí Bilabial Alveolar Alveo-palatal Plosive Voiceless p1 t Aspirated th Fricative Voiceless s ʃ Nasal Voiced m n Liquids Flap ɾ Lateral l Glide Voiced w j Velar Glottal k h 2.2 Vowels Regarding vowels, Borgman (1959a) proposes an inventory of six phonemic vowels, which aligns with the expected number of vowels among Yanomami languages. Comparatively, the inventory proposed by Ferreira et al. (2019) differs only in the high central vowel, which the authors argue is a schwa sound. Note that nasalization is also a prominent feature among the Yanomami languages. Consequently, according to Borgman’s study, most of the Ninam vowels have nasal counterparts. Table 7 shows the inventory of phonemic vowels proposed by Borgman (1959a). Table 8 presents the contrasts for the Ninam vowels. Table 7 – Phonemic vowels of Ninam of Upper Mucajaí (Borgman 1959a: 2) Front Central Back High i, ĩ ɨ u, ũ Mid e, ẽ Low o, õ a, ã 1 As in other Yanomami languages (Migliazza & Grimes 1961: 32, Goodwin 1990: 13, Ramirez 1994: 46, Lizot 2004: XVIII, Ferreira 2011: 13a, Autori 2013: 57, and Ferreira 2017:42-43), the data from Ninam of Upper Mucajaí show that there is no contrast between voiced and voiceless plosives, except for the velar plosive which was only shown in its voiceless version. In agreement with these analyses, it was decided for the use of the voiceless plosives. 17 Table 8 – Contrast between oral vowels CONTRAST IN MEDIAL POSITION CONTRAST IN FINAL POSITION i/u [hiɾa] [huɾe] ‘to deliver’ ‘heavy’ [haɾi] [haɾu] ‘to bake’ ‘early morning’ e/o [heɾe] [hoɾe] ‘wet’ ‘coward’ [haɾe] [haɾo] ‘to go out’ ‘to heal’ a/o [haɾo] [hoɾe] ‘to heal’ ‘coward’ [aka] [ako] ‘tongue’ ‘drinking canoe’ u/o [huɾe] [hoɾe] ‘heavy’ ‘coward’ [haɾu] [haɾo] ‘early morning’ ‘to heal’ i/e [hiɾa] [heɾe] ‘handle’ ‘wet’ [haɾi] [haɾe] ‘to bake’ ‘to go out’ In addition to the nasal counterparts, words in Ninam, as well as other Yanomami languages (Ramirez 1994; Ferreira 2017), exhibit the phonological process of nasal assimilation. This assimilation process, which is only blocked by plosives, results in words that only contain oral vowels or nasal vowels [hoɾe] ‘coward’ and [hõɾẽ] ‘liar’ or [yaɾo] ‘animal’ and [yãɾõ] ‘husband.’ Table 9 shows the contrast between oral and nasal vowels. Table 9 – Contrast between oral and nasal vowels i/ĩ [kiɾi] ‘be afraid’ [kĩɾĩ] ‘little bird’ u/ũ [uɾu] [ũhũ] ‘kid’ e/ẽ [ẽɾẽɾã] [eheɾa] ‘open wildly’ ‘underneath’ o/õ [hoo] [hõõ] ‘push down’ ‘jaguar’ a/ã [tʰa] [tʰã] ‘to do’ ‘word’ DEM The phonemic chart that best represents the vowels in data collected for this study is presented in Table 10. Note that, instead of the high central vowel mentioned in Table 7, the central vowel has been added. 18 Table 10 – Phonemic vowels of Ninam of Upper Mucajaí Front High i, ĩ Mid Low e, ẽ Central Back u, ũ ə ə̃ a, ã o, õ 2.3 Syllable structure The section pertaining to syllable structure in Borgman (1959a: 12) indicates that there are at least six possibilities in Ninam of Upper Mucajaí. Table 11 displays the syllable patterns identified by Borgman. Ferreira et al. (2019) mention the possibility of vowel sequences resulting in VV patterns, which contradicts Borgman’s proposal (1959a). Table 11 – Syllable patterns found in Ninam of Upper Mucajaí (Adapted from Borgman 1959a: 12) V CV CCV CCVC CVC VC /ie/ ~ /ieb/ /d͡ʒoɾĩ / /bɾĩ d͡ʒo/ /kʰõ ɡɾĩ d/ /hamna/ /naʃi ug/ [ˈi.e] ~ [ˈi.eb] [ˈd͡ʒo.ɾĩ ] [ˈbɾĩ .d͡ʒo] [kʰõ ɡl̃ĩt] [hamˈna] [naʃi ˈug~uk] ‘blood’ ‘sweat’ ‘path’ ‘Are you going?’ ‘How?’ ‘cassava drink’ The syllabic patterns observed in the recent data align with the findings reported by Borgman (1959). The most prevalent syllable structure in the data is CV. Conversely, the CCVC pattern was among the least frequent. Table 12 provides a visual representation of the identical patterns identified by Borgman (1959) and replicated in the new dataset. Table 12 – Syllable patterns in Ninam of Upper Mucajaí V CV CCV CCVC CVC VC /ako/ /ʃe/ /akleto/ /plẽmxĩĩ/ /sik/ /ug/ [a.ko] [ʃe] [ˈa.kɾe.to] [bɾẽm.ˈxĩ.ĩ] [sik] [uk] ‘drinking canoe’ ‘hit’ ‘stretch out’ ‘sun’ ‘planted’ ‘drink’ 2.4 Word and stress According to the data, Ninam words demonstrate a predictable stress pattern, primarily falling on the penultimate syllable, except for words ending in CVC syllables. In these cases, 19 the stress is placed on the last syllable. This means that stress is not a distinctive feature of Ninam of Upper Mucajaí. Therefore, no pair of words contrasts only by the syllable on which stress falls. As Borgman (1959) explained, stress does play a role at the prosodic level, though it helps to distinguish between declarative and interrogative clauses. In example (4a) the structure of the word for ‘tree’ consists of the root wii ‘tree’ and the nominal classifier =tihi ‘wood.’ The stress in this word is placed on the penultimate syllable. On the other hand, in example (4b), the nominal classifier is indexed to the verbal complex. In this case, the stress no longer falls on the nominal classifier. Still, it occurs on what is now the penultimate syllable of the word. (4) a. /wii =ˈtihi/ tree =CLN:wood ‘tree’ b. /tihi= po =la =ˈle =i/ CLN:wood= carry =CAUS =RES =?? ‘(he) carries the wood.’ 2.5 Orthography The orthography employed to present the data in subsequent sections corresponds to the system utilized in the Upper Mucajaí region. This orthographic system had its official origins in the Seminar on Orthography for the Yanomami and Macuxi languages conducted by MEVA and the Summer Institute of Linguistics (SIL) in collaboration with the National Indian Foundation (FUNAI) in 1976. Since then, there have been a few changes based on new analysis. According to the seminar's report (1976), the initial analysis treated the phone [d͡ʒ] as a separate unit from the phoneme /j/. When separated, it was treated as the phoneme /t͡s/, and its grapheme was . Subsequent studies demonstrated that the phone [d͡ʒ] was, in fact, an allophone of the phoneme /j/, and as a result, it came to be represented by the same grapheme, . Another change is related to the semivowels [w] and [j] in coda position. Originally, these were represented as and , but after the new analysis involving the allophones [d͡ʒ] and [ɲ], the phoneme /j/ came to be represented as in onset position and as in coda position. Meanwhile, the phoneme /w/ remained as in the onset and as in the coda. 20 According to Ferreira (2017), the orthography used by the Ninam of Upper Mucajaí, as well as the one used in Palimiu, differs from that used in other Yanomami languages. Nevertheless, it is worth noting that the primary distinction lies in the use of the grapheme instead of and the grapheme used for the central vowel [ə], which is represented as <è> in Ninam. Table 13 shows the phonemes of the Ninam language of the Upper Mucajaí and their corresponding graphemes. Table 13 – Current orthography system of Ninam of Upper Mucajaí VOWELS /a/ /e/ /i/ /o/ /u/ /ə/ à à à à à à <è> NASALIZATION ~ CONSONANTS /h/ /k/ /l/ ~ /ɾ/ /m/ /n/ /p/ /s/ /t/ /tʰ/ #/w/ /w/# /ʃ/ #/j/ /j/# à à à à à à

à à à à à à à à 21 3. The verb stem in Ninam of Upper Mucajaí In the languages from the Yanomami family, the verb is known for its complexity, mainly due to its morphological structure. According to Ferreira, Machado & Benfica (2019: 126), certain Yanomami languages have up to thirty-one potential slots within the verbal complex. Scholars such as Swain (1971), Goodwin (1991), Ramirez (1994), and Ferreira (2017) have approached the analysis of Yanomami verb morphology by first classifying the different types of verbs. This classification helps in determining the grammatical categories that can be cliticized to the different verb types. In her work on the Ninam spoken in the Upper Mucajaí region, Swain (1971: 19–22) presents a five-term classification of verb phrases. Swain identifies Descriptive, Position, Motion, Action-on-goal, and Equative verb stems. While maintaining the core principles of Swain’s original classification, it is necessary to make slight modifications to the terminology used to ensure better alignment with current linguistic conventions and facilitate easier cross-linguistic comparisons. In this work, descriptive, equative, and positional verbs are grouped under the label of STATIVE VERBS, and verbs of movement and action-on-goal are grouped under the label of DYNAMIC VERBS. 3.1 Stative verbs The verb stems grouped in the stative category are copular, existential, descriptive, and positional verb stems. These verbs have been grouped by the type of morphology that they host in the verbal complex. 3.1.1 Copular and existential verb stem The verb ku can be translated as ‘to be’ or ‘to exist’, thus it is used in equative and existential constructions, which place two elements in apposition or in a state of equality with each other as in example (5). In addition to this, the verb pemi indicates the non-existence of something, as in the example (6). (5) naxi =kok pè= cassava =CLN:root unsp= ‘It is cassava’, ‘the cassava exists.’ ku COP =le =PRS 22 (6) kathooxek yãa =wei ha= pemi =o =elei banana.specie tie =DEVERB HEAR= non.existence =MID.VOI =REM.PST ‘There was nothing to eat (lit. there was no hung banana).’ As observed in example (7), the morpheme ku has the allomorph ki. This variation was also identified in Yanomama by Ferreira (2017) who comments that the vowel in ku harmonizes with the vowel of the following morpheme. (7) pakam patathepata apè= ki =i fish huge UNSP= COP =?? ‘It is a huge fish.’ (CS71) 3.1.2 Descriptive In Ninam of Upper Mucajaí, it is observed that property concepts can have the modification function of an adjective or the predication function of a verb. This characteristic becomes evident when a property concept hosts verbal morphology, resulting in a descriptive verb. In Ninam, descriptive verbs define the state or condition of the subject of the sentence and usually occur in intransitive constructions. In example (8a), the term uxi ‘black’ is observed as a part of the noun phrase in a transitive clause. Its position within the noun phrase, with the ergative marker occurring after all constituent elements, conclusively establishes its affiliation with the noun phrase. Conversely, in (8b), the same morpheme is situated within the verb phrase. Its affiliation with the verb phrase becomes more evident through the use of the evidentiality marker denoted as UNSPECIFIED, which only occurs on verbal elements. The presence of verbal morphology enables its classification as a descriptive verb. (8) a. okolo uxi =in kalatha apè= wa =he dog black =ERG chicken UNSP= eat =REC.PST2 ‘The black dog ate the chicken.’ b. okolo apè= uxi =i dog UNSP= black =?? ‘The dog is black.’ Example (9b) shows the same situation with the property concept wehe ‘dry’ when used as a descriptive verb. 23 (9) a. wii =tihi poko wehe tree =CLN:wood branch dry ‘On the dry branch’ =hè =LOC b. apè= wehe =ma =po =wei UNSP= dry =CAUS =CONS =HAB ‘It makes it to stay dry.’ 3.1.3 Positional Verb stems that fall into the category of positional verbs define the position of the subject of the sentence. Usually, positional verbs require an intransitive construction, as shown in example (10a) and (10b), where the positional verbs pulu ‘lie in hammock’ and ila ‘stand’ convey the position of the subject Xakalina and the 1st singular ya= respectively. (10) a. Xakalina ape= pulu =le Jacqueline UNSP= lie.in.hammock =PRS ‘Jacqueline is lying in the hammock.’ b. ya= ila =o =i 1SG= stand =MID.VOI =IMM ‘I’m standing’ or ‘I’m about to stand.’ (CS71) 3.2 Dynamic Different perspectives have been offered regarding dynamic verbs. Ferreira (2017) highlights that dynamic verbs can be either transitive or intransitive, leading to a wide range of options. Conversely, Ramirez (1994) introduces separate categories for dynamic, transitive, and intransitive verbs, and Swain (1971) uses Motion verbs and Action-on-goal as labels for the category. In this description, I will adopt the same approach as Ferreira (2017) by understanding that dynamic verbs, regardless of their transitivity, are used to express the action or motion performed by the subject. Example (11) shows the verb hoke ‘travel’ which, in addition to being a motion verb, is intransitive and, therefore, requires only one argument as the subject. In this case, the 1st singular =ya fills that position. (11) ya= pè= hoke 1SG UNSP= travel ‘I traveled.’ (CS71) =yo =RES =heli =ITIV =he =REC.PST2 24 In example (12), the transitive verb tha ‘do, make’, which can be classified as an action, has two arguments ipa thuwen ‘my woman’ as the subject – identified by the ergative marker – and haklak=uk ‘cassava drink’ as the object. (12) ipa thuwe 1SG.POSS woman =n haklak =uk apè= tha =ERG edible.yucca =CLN:liquid UNSP= make =li =he =RES =REC.PST2 ‘My wife (lit. my woman) made the cassava drink.’ Different from research conducted for other Yanomami languages, Ninam of Upper Mucajaí does not present morphology that specifically indicates verb class, e.g. for Yanomama, Ferreira (2017: 193) identified four morphemes in present tense readings: the morpheme =a POST accompanies positional verbs, =ɨ DYN accompanies dynamic verbs, =o STV accompanies irregular verbs, and zero marking accompanies attributive verbs. However, this is not what is observed in Ninam of Upper Mucajaí. Example (13) gives data from Yanomama taken from Ferreira (2017: 188), here the morpheme =a, after the verb stem pɨrɨ 'to lie in the hammock', indicates that this is a position. (13) ɨha ya= pɨrɨ =a 1SG= lie.in.hammock =POST ‘I live there (lit. ‘I lie in the hammock there’).’ (Ferreira, 2017: 188) ANA:there In contrast, Ninam of Upper Mucajaí shows that this positional verb stem can host similar morphemes that in Yanomama are claimed to correspond to other verbal classes. Example (14) shows no other morpheme besides tense – zero marking. Example (10) repeated here as (15) shows the morpheme =i labeled IMMEDIATE. Example (16) shows the morpheme =a PERFECTIVE, and example (17) shows the morpheme =o MIDDLE VOICE. This indicates that Ninam does not have morphemes indicating verb classes. (14) Xakalina ape= pulu =le Jacqueline UNSP= lie.in.hammock =PRS ‘Jacqueline is lying in the hammock.’ 25 (15) Xakalina ape= pulu =i Jacqueline UNSP= lie.in.hammock =IMM ‘Jacqueline was about to lie in the hammock.’ (16) Naa =ye pulu =a mother =1SG.POSS lie.in.hammock =PFV ‘My mother lies in the hammock.’ (CS71) (17) Ya= pulu =o =paken 1SG= lie.in.hammock =MID.VOI =DETERM ‘I am determined to lie in the hammock.’ (CS02b) 26 4. Preverbal morphology As previously stated, Ninam of Upper Mucajaí exhibits polysynthetic characteristics. Specifically, when it comes to the verbal complex, it is noteworthy that the verbs are open on both sides to host morphological elements. Table 14 presents the grammatical categories that can be hosted on the left side of the verbal complex. These categories, namely subject and object marking, simultaneous marker, and evidentiality, are described in this section. Note that although adverbial notions can be hosted on the left side of the verbal complex, they will be described in §5.3 because their unmarked position is postverbal. Table 14 – Position for the proclitics in the verbal complex of Ninam of Upper Mucajaí SUBJECT/OBJECT SUBJECT/OBJECT SIMULTANEOUS EVIDENTIALITY ADVERBIAL 3 ya= 1SG:SUBJ/OBJ yeh= 1DL:SUBJ yam= 1PL:SUBJ ø 3SG:SUBJ/OBJ kup= 3DL:OBJ pèk= 3PL:OBJ CLN= kãi= SIM STEM apè= ~ pè UNSP ta= VIS ha= HEAR yehek= 1DL:OBJ yamak= 1PL:OBJ wa= 2SG:SUBJ/OBJ weh= 2DL:SUBJ wam= 2PL:SUBJ wehek= 2DL:OBJ wamak= 2PL:OBJ 4.1 Subject and object marking Ninam indexes pronominal markers that encode subject and object position in the verbal complex. This indexation can decrease a participant’s discursive salience (Ramirez 1994: 385). 4.1.1 Pronouns and pronominal markers Based on the corpus under consideration, pronominal reference to arguments of the clause can be done with independent pronouns and markers that cliticize to the verb. Both forms are found in Table 15. 27 PL DL SG Table 15 – Pronominal markers and Pronouns in Ninam of Upper Mucajaí Pronominal markers Pronouns 1st 2nd 3rd 1st 2nd SUBJ OBJ SUBJ OBJ SUBJ OBJ ya= wa= ø yeh = yehek = yam= yamak= weh= wehek= wam= wamak= =pu kup= =he pèk= 3rd kamiye kahowa kama kamiyehek kahowehek kamakup kamiyamak kahowamak kamapèk The parameter used to determine if the pronominal markers are indexed to the verb is the stress placement in the phonological word. However, most pronominal markers that are postulated to be indexed have only one syllable, so it becomes challenging to determine where the stress is located. Thus, the pronominal markers for dual and plural in the object position, which have two syllables, are the best data to evaluate. Following the intonation pattern in red, in Figure 4, it is noticed that the syllable hek in the noun phrase kayehek 1st dual has a higher intonation, which is the parameter used to determine word stress. Figure 4 – Spectogram and intonation of the phonological word kayehek 1st dual person. Figure 5 shows that when the pronominal marker yehek= 1st dual is indexed to the verb in yehekẽawei ‘we generally work’. The syllable hek receives what could be considered 28 as a secondary stress. Meanwhile, the higher intonation of the whole construction falls on the syllable we. Figure 5 – Spectogram and intonation of the phonological word yehekẽawei ‘we generally work.’ It was also observed that Ninam presents three subsystems to organize the pronominal marker in subject and object position within the transitive sentence. These subsystems are described in the following subsections. 4.1.1.1 Subsystem A: First/Second person acts on Third person According to the data, when the first or second person is the subject of the sentence, and the third person is the object, the subject precedes the object, as shown in Table 16. Table 16 – Sequence for subject and object marking in Subsystem A SUBJECT OBJECT Other morphemes VERB STEM ya= 1SG ø 3SG yeh= 1DL kup= 3DL yam= 1PL pèk= 3PL wa= 2SG weh= 2DL wam= 2PL 29 This sequence is illustrated in examples (18), (19) and (20) below. Example (18) shows the 1st singular marker ya= in the subject position, followed by the 3rd plural marker pèk=. (18) ai =pèk ya= pèk= miyawe =ma =ma other =PL.AN 1SG= 3PL.OBJ= smart =CAUS =REC.PST1 ‘I made others smart’ Example (19) shows the marker for the 1st dual marker yeh= followed by the 3rd plural marker pèk=. (19) inaha yeh= pèk= miyawe =ma =ithe thus 1DL.SUBJ= 3PL.OBJ= smart =CAUS =FUT ‘Thus, we (dual) will make them smart.’ Example (20) shows the pronoun for the 3rd singular in a noun phrase and the pronominal marker for the 2nd singular marker attached to the verbal complex. (20) kama wa= pè= ta =po =le 3SG 2SG= UNSP= see =CONS =PRS ‘You keep seeing him’ 4.1.1.2 Subsystem B: First/Second person acts on First/Second person Unlike other Yanomami languages, such as Ỹaroamë (Ferreira 2011: 50–52), Ninam of Upper Mucajaí does not index both subject and object when they refer to the 1st and 2nd person. In this case, Ninam only allows the indexation of one of the pronominal markers, c.f. Table 17. The other participant will be realized as a noun phrase. 30 Table 17 – Sequence for subject and object marking in Subsystem B. SUBJECT/OBJECT Other morphemes VERB STEM ya= 1SG:SUBJ/OBJ yeh= 1DL:SUBJ yam= 1PL:SUBJ yehek= 1DL:OBJ yamak= 1PL:OBJ wa= 2SG:SUBJ/OBJ weh= 2DL:SUBJ wam= 2PL:SUBJ wehek= 2DL:OBJ wamak= 2PL:OBJ In example (21b) is observed that Ninam indexes the subject. However, example (21b) is only possible when a sentence like (21a) already exists in the discourse context. As Ninam does not index both 1st and 2nd in subject and object positions, an option like (21c) constitutes an ungrammatical construction. As proposed by Albright (1965) and Swain (1971), example (21b) could even have a more passive reading, which would actually be linked to the idea of placing the object in focus since the transitivity of the verbs is not affected. This coincides with what Ramirez (1994: 385) says regarding the indexation of one of the participants in the verb to decrease that participant’s discursive salience. Examples in (22) reinforce this hypothesis concerning the 1st and 2nd person. Examples (21a), (21b), (22a), and (22b) show the grammatical forms of sentences in Ninam where either a 1st or 2nd person act on a 1st or 2nd person, as pronouns in noun phrases or as markers indexed in the verb. (21) a. kamiye =n kahowa apè= 1SG =ERG 2SG UNSP= ‘I see you’ or ‘I am seeing you.’ ta see =po =CONS b. kahowa ya= apè= ta =po =le 2SG 1SG= UNSP= see =CONS =PRS ‘I see you’ or ‘I am seeing you’ or ‘you are seen by me.’ c. *ya= wa= apè= ta =po =le =le =PRS 31 (22) a. kahowa =n kamiye apè= 2SG =ERG 1SG UNSP= ‘You see me’ or ‘You are seeing me.’ ta see =po =CONS =le =PRS b. kamiye wa= apè= ta =po =le 1SG 2SG UNSP= see =CONS =PRS ‘You see me’ or ‘You are seeing me’ or ‘I am seen by you.’ c. *wa= ya= apè= ta =po =le Since Ninam does not have a morpheme indicating focus, this discourse choice is achieved through the indexing of one of the elements and the realization of the other as a noun phrase or an indepent pronoun. If the subject is the focus of the sentence, it will carry the morpheme =n ERGATIVE and the other element will be indexed. If the focus is on the object, it will occur as a noun phrase or an independent pronoun and the subject will be expressed as a marker in the verbal complex. A third option is that both subject and object are realized as noun phrases or independent pronouns. 4.1.1.3 Subsystem C: Third person acts on First/Second/Third person A different distribution is observed when the 3rd person functions as the subject of the transitive sentence. Since the third person singular has no marking, this difference is only noticeable when it is the 3rd plural acting as the subject of the transitive sentence. Note that the marker =pu referring to the 3rd person dual subject was noted by Swain (1971), but it was not found in newer data. Table 18 shows the order for the morphemes in Subsystem C. Table 18 – Sequence for subject and object marking in Subsystem C. OBJECT Other morphemes Verb stem Other morphemes ya= 1SG yehek= 1DL yamak= 1PL SUBJECT ø 3SG =pu 3DL =he 3PL wa= 2SG wehek= 2DL wamak= 2PL ø 3SG kup= 3DL pèk= 3PL The morpheme =n ERGATIVE in examples (23), (24), and (25) indicates that these noun phrases are fulfilling the function of the subject and the morpheme =pèk(u) indicates 32 that the subject is plural. When the third person plural is the subject in a transitive sentence, the morpheme marking the subject does not occur in the preverbal position but is placed on the right side of the verbal stem, c.f. example (23), (24), and (25). (23) missionário =pèku =n ya= miyawe =ma =le =ma missionary =3PL =ERG 1SG.OBJ= smart =CAUS =RES =REC.PST1 =he =3PL.SUBJ ‘The missionaries made me smart.’ (24) ũhu that =thehe =when ai other missionário =pèku =n yehek= missionary =3PL =ERG 1DL.OBJ= thẽlema =ma =he help =REC.PST1 =3PL.SUBJ ‘In that time, other missionaries took care of us (dual).’ (25) Pata =pèku =n wake hipo =ha =lã =elai old =PL.AN =ERG fire own =HEAR =NEG =REM.PST2:3PL.SUBJ ‘The elders had no fire.’ As opposed to the 3rd plural, examples (26) and (27) illustrate that there is no morpheme indexed to the verb indicating 3rd singular. In example (26), the noun phrase is composed of the 3rd singular pronoun kama followed by the ergative =n indicating that it is the subject of the sentence. However, the verb does not show any morpheme to indicate 3rd singular. In example (27) the 1st plural pronoun kamiyamak is the object as it is not followed by the ergative marker. In this instance, it becomes evident that the absence of marking indicates 3rd singular. (26) Kama =n wehek= apè= ta =po =le 3SG =ERG 2PL.OBJ= UNSP= see =CONS =PRS ‘He keeps seeing you (pl).’ (27) Kamiyamak apè= ta =po =le 1PL UNSP= see =CONS =PRS ‘He keeps seeing us.’ 33 4.1.2 Noun classifiers In Ninam, the classifiers can be used anaphorically to refer to a previously mentioned referent. In such cases it cliticizes to the verbal complex. In example (28), the noun classifier tihi ‘wood,’ accompanies the noun wii ‘tree.’ Further on, the classifier is cliticizied to the verbal complex, indicating a link with the previous sentence. (28) wãlo man =n wii =ERG tree =tihi te =i, ũhu =thehe =CLN:wood grab =?? that =when tihi= po =la =le =i CLN:wood= carry =PLU =RES =?? ‘The man gets the firewood, then carries the firewood.’ A way to verify that the clitic =tihi ‘wood’ has been indexed to the verb is by studying the placement of stress in the word. As mentioned in §2.4, stress is normally placed on the word’s penultimate syllable, except for specific cases as CVC syllables at the word ending. In Figure 6, intonation is marked by the red line. In the word wiitihi ‘wood of the tree’ the stress, the rise in pitch, falls on the syllable ti. Figure 6 – Spectrogram and intonation of the phonological word wiitihi ‘wood of the tree.’ 34 As for the subsequent sentence, following the red line in Figure 7, it is noted that in tihipolalei the stress falls on the syllable le. This shows that the noun classifier tihi= ‘wood’ now belongs to a different chain of morphemes. Figure 7 – Spectrogram and intonation of the phonological word tihipolalei ‘about to carry the wood.’ The anaphoric functions of classifiers are easier to see in longer texts. In (29a), the narrator introduces the noun haklak ‘edible yucca’ with two classifiers, first with =kok ‘root’ and then with =uk ‘liquid.’ Then when the narrator presents a sequence of actions, he only uses the classifiers to refer to the noun mentioned, kok ‘root’ in (29b-f) and uk ‘liquid,’ in (29g). Note that (29g) shows it cliticized to the verb but before the evidential marker, which is the expected position for an object marker. (29) a. thuwe woman ha= SEQ= =n =ERG hutu farm =hè =LOC haklak edible.yucca =kok =CLN:root huku pull.out =li =RES =n =after haklak edible.yucca =uk =CLN:liquid apè= tha =wei. UNSP= do =HAB ‘After the woman pulls the cassava in the field, she makes the cassava drink.’ b. Hutu =hè kok= te =wei, farm =LOC CLN:root= grab =HAB ‘She grabs the root in the farm.’ 35 c. Yãno =hè kok= kãi= house =LOC CLN:root= SIM= ‘She returns home with the root,’ d. kok= hõxi =la CLN:root= peel =PLU ‘She peels the root,’ =le =RES e. kok= yalu =le CLN:root= wash =RES ‘she washes the root,’ =i =?? f. kope =i return =?? =i =?? kok= tu =ke =n waiha CLN:root= cook =INCEP =after after.waiting ‘after cooking, after waiting,’ g. kama =n uk= apè= tha =wei 3SG =ERG CLN:liquid= UNSP= do =HAB ‘She makes the drink.’ 4.2 Simultaneous In other Yanomami languages, the morpheme kãi= has been categorized as both COMITATIVE (Ramirez 1994: 156) and SIMULTANEOUS (Lizot 2004: 144). This morpheme is employed to convey the concept of accompaniment and to express the simultaneity of events or situations. The translations provided by the Language Resource Person (LRP) revolve around the words ‘together’ and ‘with,’ as in example (30), while also encompassing the notion of ‘at the same time,’ as illustrated in example (31). (30) Xosimaa Josimar =n =ERG yehek= 1DL.OBJ= kãi= SIM= lele run =a =PLU =ma =REC.PST1 =hèm =DIR ‘Josimar and we (dual) ran to the store’ Or ‘Josimar took us with him to the stores’ loxa store 36 (31) ãluhu fermented.cassava =uk =CLN:liquid kãi= SIM= apè= UNSP= pa place.on.the.ground =o =wei =MID.VOI =HAB ‘The fermented cassava drink (is already) placed on the ground at the same time.’ 4.3 Evidentiality According to Aikhenvald and Dixon (1998), evidentiality as a grammatical category is a common feature among the languages of the Amazonia area. The data suggest that Ninam of Upper Mucajaí has a three-term system composed of VISUAL, HEARSAY, and UNSPECIFIED.2 Unlike the other Yanomami evidentiality systems that have been documented (Borgman 1990, Gómez Goodwin 1991, Ramirez 1994, and Ferreira 2017), this is the only language where the unmarked position for the evidentials is a preverbal position. 4.3.1 Visual The evidential marker ta= ~ tak= labeled as VISUAL refers to information obtained through seeing and it extends to direct observation, participation, control, observable facts, and certainty (Aikhenvald 2003: 13). In Ninam, the evidentiality marker ta= ~ tak= VISUAL is a grammaticalized form of the verb ta ~ taa ‘see’. Both can even occur in the same sentence as shown in example (32). (32) lilihi =hèm thuwe =n paali ta= taa =li =i forest =DIR woman =ERG bare.faced.curasow VIS= see =RES =PST ‘The woman saw the curassow in the forest.’ Albright (1970:1) suggests that the evidential ta= VISUAL is related to the narrator's involvement in the narrated event. Example (33) indicates this possibility; however, other examples do not indicate anything directly related to the narrator's involvement, e.g. example in (34) indicates an event in which the narrator seems to have no involvement other than mere observation. 2 Albright (1970) described a five-term system for Ninam of Upper Mucajaí. This system included the labels and PARTICIPATED. The newer data did not contain these markers. It is assumed that the evidential system has been reduced. DEDUCED 37 (33) inaha ku thus COP yalo because Teus e thã God 3SG.POSS word =elihè =LOC ya= 1SG= kẽa =wei sĩuthai ta= ku =o =lei work =NMLZ a.little.bit VIS= COP =MID.VOI =REM.PST ‘Thus, I did a little bit of work on the Word of God.’ (34) Mana =hèm naa =ye ta= lãamo =yo =i river =DIR mother =1SG.POSS VIS= be.sick =RES =PST ‘In the direction of the river, your mom got sick.’ According to the data, the tak= variation is exclusively used for constructions whose readings are in the present tense. Leaving the ta= form for constructions whose readings are in the past tense. This is observed in the data in (35) since they differ only in the morpheme indicating tense. The same occurs in (36). (35) a. kama kawik tak= ku =le 3SG beard VIS= COP =PRS ‘His beard exists.’ b. kama kawik ta= ku =elei 3SG beard VIS= COP =REM.PST ‘His beard existed.’ (36) a. Yãno kami =hèm matohip tak= ala house inside =DIR things VIS= place.on.top.of ‘The possessions are on a shelf inside the house’ =le =PRS b. Yãno kami =hèm matohip ta= ala house inside =DIR things VIS= place.on.top.of ‘The possessions were on a shelf inside the house’ =lei =REM.PST 4.3.2 Hearsay Another evidential marker found in the Ninam data is ha= ~ han= labeled HEARSAY. This evidential refers to information acquired through a third party. According to Albright (1970), the hearsay evidential could also indicate the degree of involvement of the third party. However, since most of the data with this marker are mythical narratives it is impossible to verify the degree of involvement of the person reporting. Examples (37) and (38) both present the evidential ha= HEARSAY on the left side of the verb. 38 (37) pata =pèk =un wake hipo ha= =lã =elahei old =PL.AN =ERG fire own HEAR= =NEG =REM.PST:PL ‘The elders had no fire.’ (38) ninam =un yalimi huxmili ha= taa =elei person =ERG spider.monkey dead HEAR= see =REM.PST ‘A person saw the dead spider monkey.’ Similarly to the VISUAL evidential, the morpheme indicating heard evidence has a variation. According to Albright's (1970) data, elicited again for this research, the variation han= is used in the context of direct quotation as in the examples (39), (40), and (41). It cooccurs with the morphemes =le PRESENT, =he RECENT PAST 2, and the morpheme =i. (39) Naa =ho han= kili =le Mom =2sg.poss HEAR= be.scared =PRS ‘Your mother says that she is scared.’ (SA70) (40) ai them =ũhu xama yãhi han= wa =li =he other day =DEM tapir meat HEAR= eat =RES =REC.PST2 ‘He says that he ate tapir’s meat the other day.’ (SA70) (41) kama han= thomo =li =i 3SG HEAR= steal =RES =PST ‘He says that it was stolen.’ 4.3.3 Unspecified Ninam expresses non-personal knowledge or unspecified source with the morpheme apè= UNSPECIFIED. The morpheme apè= ~ pè= is also the default evidential used when eliciting isolated sentences. According to examples in (11), here repeated as (42), (21), and (27), the morpheme can be used when the speaker is involved in the event being described. In these cases, the meaning of the morpheme is not non-personal knowledge but rather uspecified source. This morpheme was also found in two texts where the male speaker describes activities carried out only by the women in the village. As this evidential could express nonpersonal knowledge, it makes sense, in this case, for a man to use it, c.f. examples (44) and (45). (43) ya= pè= hoke 1SG UNSP= travel ‘I traveled.’ (CS71) =yo =RES =heli =ITIV =he =REC.PST2 39 (44) haklak =uk apè= tha =wei edible.yucca =CLN:liquid UNSP= do =HAB ‘(She) makes the cassava drink.’ (45) nax =kok apè= hõxi =ma =wei poaka =un wild.yucca =CLN:root UNSP= peel =CAUS =HAB machete =INSTR ‘(She) peels the cassava with a machete.’ (46) kamiyeheku =n wehek= apè= taa =po =le 1DL =ERG 2DL UNSP= see =CONS =PRS ‘We (dual) see you (dual).’ 4.3.4 Zero marking According to the data, the evidentials in Ninam appear to be obligatory in sentence formation, except for sentences with RECENT PAST 1=ma, c.f. examples (47) and (48). According to Ferreira (2017: 619) this also occurs in Yanomama. (47) ũhu=thehe ai DEM=when other missionário =pèk =un yehek= thẽlema missionary =PL =ERG 1DL.OBJ= help =ma =he =REC.PST1 =3PL.SUBJ ‘In that time, other missionaries took care of us (dual)’ (48) Xosimaa =n yehek= thẽlema =ma Josimar =ERG 1DL.OBJ= help =REC.PST1 ‘Josimar helped us.’ It was also observed that the data collected in the future tense did not present any morpheme coding evidentials, c.f. examples (49) and (50). (49) (50) haximi.watu =wei pata~pata rest =NMLZ big~AUG ‘There will be a huge rest.’ ku COP =o =MID.VOI yeh= pèk= miyawe 1DL= 3PL= smart ‘We (dual) will make them smart.’ =ma =CAUS =ithe =FUT =ithe =FUT 40 This makes sense in relation to the future tense since there is no source of information for something that has not occurred. This also indicates that the UNSPECIFIED evidential does not have the sense of ‘probability’ unlike in some other Yanomami languages (Infante 2023), e.g. Sanuma has at least one evidential that encodes probability and it is used in constructions with future tense. The data from Ninam leads to the assumption that the evidentials only occur in present and past tense. 4.3.5 Note on evidentials and adverbial notions Based on the examples presented in this section, the evidentials in Ninam typically occur on the left side of the verb stem. However, examples (51) and (52) demonstrate an alternative placement of evidentials on the right side of the verb stem. In example (51), the evidential =ha HEARSAY occurs between the adverbial notions =kõ ‘again’ and =lã NEGATION, after the verb ĩka ‘to smile’. Example (52) shows the evidential =pè UNSPECIFIED following the copula and preceding the adverbial. These instances indicate that the position of evidentials can change in response to the presence of adverbial notions within the verbal complex. (51) ĩka =kõ =ha =lã =elei smile =again =HEAR =NEG =REM.PST ‘Again, (the alligator) did not smile.’ (52) nini =pè =haiki =le be.in.pain =UNSP =certainly =PRS ‘He is in pain.’ 41 5. Postverbal morphology As previously mentioned, the verbs in Ninam are open on both sides to host morphological elements. Table 19 presents the grammatical categories that can be hosted on the right side of the verbal complex. These categories, namely valency change, adverbial notions, aspect markers, directionals, locatives, tense and mood markers, and subject markers, are described in this section. Table 19 – Position for the enclitics in the verbal complex of Ninam of Upper Mucajaí VALENCY ASP/ DUR DIR/ CHANGE STEM ADV 2 EVID ADV1 MID.VOI/ LEVEL PLANE 1 2 3 =a =ma CAUS apè= UNSP =layo RES =heli ITIV =mo REFL ta= VIS =le RES =ima VEN =yo RECIP ha= HEAR =ke INCEP =a PFV =o MID.VOI =po CONS =la ‘horizontal’ TENSE/ ASP2/ MOOD =lei REM.PST =ma REC.PST1 =he REC.PST2 =lexi PRE.HOD =xi HOD =le PRS =i IMM =ithe FUT 3SUBJ =he(i) 3PL =pu 3DL =li ‘up’ =kili ‘down’ =ti ‘near’ =wei HAB/INF =lo ‘around’ =i IMP =paken DETERM =xo COMIT 5.1 Pluractional =a The morpheme =a has been described as ‘turn into’ (Barker 1963), ‘stative focus’ (Tracy 1964: 41 and Swain 1971), ‘minus reflexive’3 (Migliazza 1972: 135), ‘transitivizer’ and ‘perfective’ (Ramírez 1997: 428), ‘stative perfective’ and ‘repeated action’ (Lizot 2004: 1– 2), and ‘positional,’ ‘perfective,’ and ‘distributive’ marker (Ferreira 2017: 326–329). Considering that the verbal complex in Ninam is very consistent in relation to the positions occupied by the morphemes, it is possible that we are dealing with the same form for at least two different morphemes. 3 Migliazza (1972) proposes that the morphemes =o and =a are types of reflexives. Thinking in terms of duality, the morpheme =o is plus reflexive, and the morpheme =a is minus reflexive. 42 The following examples show that there are at least two different positions for the morpheme =a; one slot before the valency change morpheme, as in example (53) and one after the adverbials, as in examples (54) and (55). In addition to this, example (53) also has the causative and the evidential, which ensures that it is a completely different position than the one used in (54) and (55). (53) lau =a =ma =pè =kõ =o =wei dance =PLU =CAUS UNSP= =again =MID.VOI =HAB ‘(all) dance again.’ (54) ya= kõ =kõ =palu =a 1SG= come.back =again =first =PFV ‘I come back first again.’ (SA65) (55) ya= pulu =haiki =a 1SG= lie.in.hammock =certainly =PFV ‘I’m certainly lying in the hammock’ (CS02b) In the first position, right after the verb stem, the morpheme =a~=la may indicate that the action performed is repeated (Lizot 2004: 2 and Ferreira 2017: 327). Example (56) refers to a party where Yanomami flutes are played. Although the free translation suggests 'to play', the verb used means ‘to blow’, so it would make sense to describe a repeated action if the purpose is to play the flute. This could also explain the use of the morpheme =a with the verb lau ‘dance’ in example (53) above. (56) ai =pèk other =3PL dois two =un tokalusi ai =ERG horn other apè= hola UNSP= blow =a =PLU =pèk =3PL =un thõlema inaha =ERG flute thus =we =hei =HAB =3PL.SUBJ 'Thus, some (play) the horn and others the flute, two (people) play.’ Ferreira’s (2017: 326) approach also mentions that this morpheme could also be considered a DISTRIBUTIVE because it accompanies distribution verbs where several objects or beneficiaries are involved. The sentences in (57) and (58) with the verb hipi 'to give' have the idea that there are several beneficiaries. The same is true in example (59), where several people will receive the word that will be transmitted. 43 (57) ũhu DEM =thehe kẽa =when work =wei thuwe =n =NMLZ woman =ERG haklak =uk apè= cassava =CLN:liquid UNSP= hipi =a =wei kẽa =wei =pèk =hèm give =PLU =HAB work =NMLZ =PL.AN =DIR 'So, the working woman give the drink to the workers.' (58) ãluhu =uk yalami apè= hipi =a =o =wei cassava =CLN:liquid much UNSP= give =PLU =MID.VOI =HAB ‘A lot of fermented cassava drink is given’ (59) ninam =pèk =hèm yeh= thã= hayi =a =ma =ithe person =PL.AN =DIR 1DL.SUBJ= word= pass =PLU =CAUS =FUT ‘We will pass the word to the Ninam.’ Examples (57) and (58) contrast with example (60) where there is only one object or beneficiary in question, and it is noted that the morpheme =a is absent. (60) wãlima =n apè= hipi =ki =he friend =ERG UNSP= give =INCEP =REC.PST2 ‘My friend gave it.’ Despite all the analysis made by other authors, the decision in this analysis is to group both the ITERATIVE and the DISTRIBUTIVE functions under the label PLURACTIONAL. 5.2 Valency change Three markers for valency change were observed in the Upper Mucajaí Ninam data: =ma CAUSATIVE, =mo REFLEXIVE, =yo RECIPROCAL. Concerning their position in the verbal complex, it is observed that these morphemes are always found between the verb and the adverbials. 5.2.1 Causative =ma The morpheme =ma is used to increase the valency of verbs. When used, it implies that an agent was responsible for causing what is being described. When comparing the stative sentence in (61) with the corresponding causative sentence in (62), it is observed that the element that was the subject being described in example (61) becomes the object in example (62). 44 (61) pèk= miyawe 3PL.SUBJ= be.smart ‘They got smart.’ =yo =RES =ma =REC.PST1 (62) ya= pèk= miyawe =ma 1SG= 3PL.OBJ= be.smart =CAUS ‘I made them to get smart.’ =le =ma =RES =REC.PST1 Similarly, in (63) the dynamic verb ke ‘to fall’, an intransitive verb, hosts the morpheme =ma CAUS and increases its valency resulting in the addition of a participant, in example (64). Note that in Ninam, the 3rd singular is not indexed to the verb, c.f. §2.1.3. Therefore, both arguments are not overtly expressed. (63) apè= ke =wei UNSP= fall =HAB ‘It falls.’ (64) apè= ke =ma =wei UNSP= fall =CAUS =HAB ‘He throws it.’ With the transitive verb wa 'to eat', only two arguments are naturally expected; in example (65), naahon ‘his mother’ is indicated as the subject through the ergative marker, and yalaka ‘fish’ is the object. When the causative =ma is added, a participant is also added; in example (66) naahon is still the subject, yalaka is the direct object, and ulu ‘kid’ is the indirect object, indicated using the directional =elihè. (65) naa =ho =n yalaka mother =2SG.POSS =ERG fish ‘Your mother keeps eating fish.’ (66) ulu kid =elihè =DIR apè= wa UNSP= eat naa =ho =n mother =2SG.POSS =ERG =ma =po =le =CAUS =CONS =PRS ‘Your mother makes the kid to keep eating fish.’ =po =le =CONS =PRS yalaka apè= wa fish UNSP= eat 45 5.2.2 Reflexive =mo In sentences where the subject performs the action on itself, the Ninam speaker uses the morpheme =mo, labeled as REFLEXIVE. Examples in this subsection belong to a narrative that explains the process of the Feast of the Dead, a typical Yanomami festival. Among the things they do, the narrator explains that they rub their skin with black ink from a tree. Although the verb stem hilika ‘to rub’ in example (67) requires a subject and an object, it is observed that there is not another explicit participant, besides the 3rd plural pèk=. In this case, the morpheme =mo indicates that the subject acts upon itself. Example (67) can be contrasted with example (68) where the verb hilika does not host the reflexive and the sentence indicates what is being rubbed. (67) pèk= apè= hilika 3PL.SUBJ= UNSP= rub ‘They rub themselves.’ =mo =REFL =wei =HAB (68) pelehe =hèm popoxi head =LOC dust ‘(He) will rub dust in the head.’ hilika rub =ithe =FUT In sentences with intransitive verbs, the Ninam speaker uses the causative =ma to increase the valency of the verb, as previously shown in (64), and can then use the reflexive to indicate that the subject does the action to him/herself. In example (8b) repeated here as (69), there is no causative marker so there is no increase in valency. Consequently, the reflexive morpheme could not be present either. In contrast, example (70), which also contains the descriptive verb uxi 'to be black', has the morpheme =ma CAUSATIVE. By increasing the valency, the verb now has two arguments: someone (argument A) causing another participant (argument B) to be black. Nevertheless, as the morpheme =mo REFLEXIVE occurs after the causative, this indicates that arguments A and B are the same person. (69) okolo apè= uxi dog UNSP= black ‘The dog is black.’ =i =?? (70) pèk= apè= ux =ma =mo =wei 3PL.SUBJ= INFR= be.black =CAUS =REFL =HAB ‘They cause themselves to be black.’ 46 5.2.3 Reciprocal =yo As opposed to the REFLEXIVE, where the agent acts upon himself, the RECIPROCAL conveys the idea that an action is carried out by two or more participants on each other. In example (71) the subject is realized by a 1st plural pronominal marker and in (72) the subject is realized by a 2nd plural pronominal marker. In both cases, the morpheme =yo indicates that the action is carried out between each of them. (71) yamak= apè= thẽlema =yo =le 1PL.SUBJ= UNSP= help =RECIP =PRS ‘We help each other.’ (72) xikawe =n wamak= apè= yã =yo =he arrow =INST 2PL.SUBJ= UNSP= shoot =RECIP =REC.PST2 ‘You were shooting each other with arrows.’ In the same text on the Feast of the Dead, it is observed that the speaker of Ninam uses the REFLEXIVE and RECIPROCAL morphemes to establish the difference between actions that he performs on himself and actions that he exchanges with others. In example (73) the informant uses the reciprocal =yo followed by the causative =ma. (73) ãhi =ki =n muk apè= hilika =yo =ma =wei muddy =PL.IN =INSTR face UNSP= rub =RECIP =CAUS =HAB ‘(they) cause to rub mud in each other's faces.’ Finally, in the same text, we observe that the Ninam speaker uses both reflexive and reciprocal morphemes in the same sentence, c.f. example (74). As in other examples, the causative =ma also occurs in this sentence. When comparing the data, it is observed that morphemes related to valency change do not occur in a fixed position when they occur together. It could be hypothesized that there is a large slot for valency change morphemes containing three positions where the three morphemes are organized. (74) thoothotho =hèm kõmoli =pèk apè= xai =mo =yo =ma =wei liana =LOC all =PL UNSP= drag =REFL =RECIP =CAUS =HAB ‘Everyone pushes themselves and each other on the liana.’ 47 5.3 Adverbials The adverbials occur with all verb types. Some of the adverbials that make up this category are =kõ ‘again’, =heti ‘also’, =haiki ‘certainly’, =lã NEG, =yalami ‘much’, and =palu ‘much’. It is necessary to note that when Ninam speakers use an adverbial in the verbal complex, the evidential – whose unmarked position is preverbal, c.f. example (75) – moves to the right side of the verbal complex; specifically it moves to a position immediately to the left of the adverbial. This occurs in all cases where there is an evidential marker and an adverbial, c.f. examples (76) and (77). (75) ya= pè= hu =le 1SG= UNSP= move =PRS ‘I am going.’ (76) waye =ha =yalami =o =elei be.angry =HEAR =many =MID.VOI =REM.PST ‘He was very angry.’ (77) ya= kẽa =pè =kõ =o =le 1SG= work =UNSP =again =MID.VOI =PRS ‘I am working again.’ In addition to the alternate position of the evidentials, the combination of adverbial notions brings about other changes. When the Ninam speaker needs to use two adverbials, one more slot opens up on the right side of the verb. This pattern was only observed with the combination of the adverbial =kõ(o) ‘again’ and the negative adverbial =lã, shown in example (51), here repeated as (78). In these cases, the evidential marking occurs between the two adverbials. (78) weli ĩka =kõ =ha =lã =elei alligator smile =again =HEAR =NEG =REM.PST ‘Again, the alligator did not smile.’ If the above example is compared with examples (76), and especially with (77) it can be noted that the primary position for adverbials is on the right side after the evidentials. But, 48 in the presence of two adverbials, a new slot is necessary. Therefore, the positions can be labeled as ADVERBIAL 2 and ADVERBIAL 1. The rarest case is observed in example (79), where the adverbial moved to a preverbal position next to the evidentiality morpheme. Considering that most adverbials occur on the right side, including the adverbial =kõ(o) ‘again’ as shown in example (77), it is understood that this is a marked position. (79) ha= kõ= le =heli =lei HEAR= again= run =ITIV =REM.PST ‘(he) ran away again.’ 5.4 Aspect marking On the right side of the verbal complex, after the valency change markers and adverbials, there is a slot where the morphemes =layo ~ =yo, =le ~=li ~=lu, =ke ~=ki, =a, and =po occur. This slot has been labeled as ASPECT. Note that the morpheme =o also occurs in this slot; however it is addressed in §0. 5.4.1 Resultative For the resultative aspect there are two forms that are in complementary distribution, the form =layo ~ =yo and the form =le ~ =li. They convey the idea that something changed state and became what is being described by the verb, or that process reached its final stage. The morpheme =layo ~ =yo occurs with intransitive verbs and the morpheme =le ~ =li occurs with transitive verbs. 5.4.1.1 The resultative =layo ~ =yo As shown in the examples in (80) to (84), the variation between =layo ~ =yo is not phonological, as they occur in similar phonological environments. Thus, the choice is considered to be lexical, meaning that the form =layo occurs with a set of verbs, and the form =yo occurs with a different set of verbs. (80) a. Petlu ha= Peter HEAR= ‘Peter stood up.’ ila =layo stand =RES =lei =REM.PST 49 b. Ha= HEAR= noma =yo =lei be.dizzy =RES =REM.PST ‘It got dizzy’ (81) (82) (83) (84) a. Ĩkali mu ha= yahathe =layo with.strenght head HEAR= break =RES ‘With strength, the head broke.’ =lei. =REM.PST b. Naxi kok ha= cassava =CLN:root UNSP= ‘The cassava got wet.’ =lei =REM.PST hele be.wet =yo =RES a. Xama =elihè ya= pè= kili Tapir =POST 1SG= UNSP= be.scared ‘I got scared of the tapir.’ =layo =RES =he =REC.PST2 b. Mene =e =pèk apè xuti =yo mother =3SG.POSS =3PL UNSP= be.itchy =RES ‘His mother and her sisters (lit. his mothers) are itching.’ =he =REC.PST2 a. Xama apè= lothotho =layo tapir UNSP= be.strong =RES ‘The tapir got strong.’ (CS02) =he =REC.PST2 b. Xama =pek apè= tapir =PL UNSP= ‘The tapir got sick.’ (CS02) =yo =RES a. ha= yãnuku =layo HEAR= be.calmed =RES ‘He got calmed.’ lãamo be.sick =he =REC.PST2 =lei =REM.PST b. yehek = ta= putu =yo 1DL.SUBJ= VIS= be.satisfied =RES ‘We (dual) got satisfied.’ =lei =REM.PST In addition to the multiple descriptive verbs shown in the examples above, the morpheme =layo ~ =yo also occurs with positional and dynamic verbs in intransitive constructions. The examples (85) to (87) contain the verbs ila ‘to stand’, ke 'to fall', and kẽa 50 'to work'. In these examples, the morpheme =layo ~ =yo also conveys the idea of the final stage. (85) ya= pè= ila 1SG= UNSP= stand ‘I stood (from being sit).’ (86) yalimi =n pihi =kãi ha= spider.monkey =ERG mind =SIM HEAR= 'The spider monkey fell along (with the nut).' (87) ipa 1SG.P ilihi forest =layo =RES =he = REC.PST2 ke fell =yo =RES =lei =REM.PST =hè sete ano ya= kẽa =hela =LOC seven year 1SG= work =ITIV OSS =yo =ma =RES =REC.PST1 ‘I worked back home (lit. in my forest) for seven years.' Note that the allomorph =yo is homophone to the RECIPROCAL marker =yo. As the morpheme =layo ~ =yo is associated with intransitive constructions, the RECIPROCAL =yo will never co-occur with the aspect =yo. 5.4.1.2 The resultative =le ~ =li ~ =lu The morpheme =le ~ =li ~ =lu is in complementary distribution with the morpheme =layo ~ =yo, and conveys to the idea of a final stage. It occurs with transitive constructions, as presented in examples (88) and (89), which show the transitive verbs hĩli 'to hear' and naka ‘buy’ followed by the resultative =le. (88) xapoli shaman =pèku =PL =n =ERG thã word =he =ithe =3PL.SUBJ =FUT ‘The shamans will want to hear the word' hĩli hear ximo want =le =RES 51 (89) ũhu DEM =thehe ipa sapato cento e quarenta =when 1SG.POSS shoe one.hundred.and.forty =a =wei ya= naka =le =?? =HAB 1SG= buy =RES ‘Then I bought my shoe that was 140 reais.' reais reais ku COP =ma = REC.PST1 The morpheme =le ~ =li ~ =lu also occurs with verbs whose valency has been increased by the morpheme =ma CAUSATIVE. Examples in (90), previously presented as (61) and (62), and examples in (91) feature the verbs miyawe 'to be smart' and hu 'to move', which naturally occur in intransitive constructions. In (90a) and (91a), the verbs are followed by the aspect =layo ~ =yo which only accompanies intransitive constructions. In contrast, in examples (90b) and (91b), whose valency have been increased, the resultative aspect presents the form =le, confirming that it only accompanies transitive constructions. (90) (91) a. pèk= miyawe 3PL.SUBJ= be.smart ‘They got smart.’ =yo =RES =ma =REC.PST1 b. ya= pèk= miyawe 1SG= 3PL.OBJ= be.smart ‘I made them to get smart.’ =ma =CAUS a. ya= apè= 1SG= UNSP= ‘I traveled.’ hu move =layo =RES =he =REC.PST2 b. secretaria secretary =pèku =PL =n =ERG ya= 1SG.OBJ= theli people =le =RES =ma =REC.PST1 ha go =ma =CAUS =le =ma =he =RES = REC.PST =3PL.SUBJ 'The people from the secretary's office made me leave.' The morpheme =le has the allomorph =li that occurs before the past tense marker =he ~ =ha, as shown in examples (12), here repeated as (92), and example (93). It is observed that this allomorph occurs through a dissimilation process that does not allow a sequence of morphemes with the same vowel. 52 (92) ipa thuwe 1SG.POSS woman =n =hakla =uk apè= tha =ERG edible.yucca =CLN:liquid UNSP= do =li =he =RES =REC.PST2 ‘My wife (lit. my woman) made a cassava drink.’ (93) sĩuthai apè= ta =le =ha =hei ai =pèku =n a.little.bit UNSP= see =RES =REC.PST2 =3PL.SUBJ other =PL =ERG ‘Others saw a little bit.’ 5.4.2 Inceptive =ke ~ =ki The morpheme =ke has been labeled as ‘change of position’ (Tracy 1964, Albright 1970: 7, Swain 1971), ‘perfective’ (Migliazza 1972: 140-143; Lizot 2004: 161; and Ferreira 2017: 310), ‘beginning of a new action’ (Borgman 1990: 186–192), and ‘ingressive aspect’ (Goodwin 1991: 98). In this work, the morpheme =ke, and its allomorph =ki, has been classified as the INCEPTIVE aspect. The study of this morpheme should be viewed in contrast to morphemes occurring in the same position. Thus, this morpheme contrasts with the ideas of (a) actions in the final stage marked with the RESULTATIVE, and (b) actions that remain in progress marked with the CONSERVATIVE. In general, the morpheme =ke indicates a point between the beginning of the action and an end that is not the final stage. The morpheme =ke highlights the beginning of the actions in examples (94a) and (94b). (94) a. Popa father =ho ha= =2SG.POSS SEQ= putu =yo =n, be.satisfied =RES =after apè= UNSP= pulu =ki = he lie.in.hammock =INCEP =REC.PST2 ‘After his father was satisfied, he lay down in the hammock.’ (CS02) b. Xama tapir ha= SEQ= yã shoot =la =CAUS =le =n, =RES =after naa mother =ye =n xama yãhi kãa =ke =ithe =1SG.POSS =ERG tapir meat cook =INCEP =FUT ‘After shooting the tapir, my mother will cook the meat of the tapir’ (CS02) 53 Examples in (95) contrast in the same way. The LRP says that he recorded fifty percent of the video in one week and the other fifty percent in another week. Finally, that will be one hundred percent. By using the =ke in (95a) and (95b), the Ninam speaker conveys the idea that the process of the recording is half-way done, and that the remaining will be initiated. This will lead to the result, indicated with the morpheme =layo. Thus, the morpheme =ke occurs indicating that the action is somewhere between the beginning and prior to the expected end. (95) a. ciquenta fifty porcento =elihè percent =LOC video video yeh= pè= 1DL.SUBJ= UNSP= tha la =ma =ki =he give.birth =CAUS =INCEP =REC.PST1 ‘We have created (lit.give birth) fifty percent of the video.’ b. ai other semana week =hèm =LOC ai other ciquenta fifty ku =kõ tha la =ke =ti =again give.birth =INCEP =near ‘The other week was again the birth another fifty percent.’ COP c. ũhu DEM =thehe =when cem porcento one.hundred percent ku COP porcento percent =ma =REC.PST1 =layo =RES =ithe =FUT ‘Then, it will be one hundred percent.’ 5.4.2.1 Note on the morpheme =ke ~ =ki It has also been noted that the morpheme =ke is the only aspect that occurs with constructions whose verbs indicate transfer, e.g. hila 'to deliver' and hipi 'to give.' Ferreira (2017: 314) indicates that the morpheme =ke adds the idea of ‘giving.’ He also mentions that the alternance between the aspects =ke and =le encodes the direction of the transfer in Yanomama; however, examples in (96) do not support this hypothesis in Ninam because the aspect =ke remains when the recipient changes. 54 (96) a. kahowa =n yalaka ya= 2SG =ERG fish 1SG= ‘You gave me fish.’ pè= hipi UNSP= give =ki =INCEP =he =REC.PST2 b. kamiye =n yalaka wa= 1SG =ERG fish 2SG.OBJ= ‘I gave you fish.’ pè= UNSP= hipi give =ki =INCEP =he =REC.PST2 c. kamiye =n yalaka pè= 1SG =ERG fish UNSP= ‘I gave him fish.’ hipi give =ki =he =INCEP =REC.PST2 5.4.3 Conservative =po In agreement with other work, the morpheme =po has been labeled as CONSERVATIVE because it can be interpreted as conserving, maintaining, or keeping something in a state, position, or motion, which conveys the same meaning described by other linguists such as Ramirez (1994: 252), Lizot (2004: 316), and Ferreira (2017: 284). Although it has been suggested that the conservative only occurs with descriptive verbs (Lizot 2004) or transitive verbs (Ferreira 2017), the data from Ninam suggest that this morpheme can occur with both stative and dynamic verbs. Example (97) has the CONSERVATIVE with the transitive verb tha 'to put'. In this case, the idea is to keep the cassava bread placed on the shelf. (97) kama 3SG =n nãxi =hi =k =ERG cassava =CLN:dough =PL.IN =wei =hè apè= tha =po =NMLZ =LOC UNSP= put =CONS 'She keeps the cassava breads on the shelf.' ala place.on.top.of =o =MID.VOI =wei =HAB Example (98) shows the morpheme =po with the descriptive verb wehe 'to be dry'. In this case, it indicates that the leaves are kept dry. 55 (98) maa water ke fall =yo =RES =n =after maha kii CONCS henahi temporarily.house ku haiki certainly =a =PFV =yalo CNJ.EXPLV apè= UNSP= wehe be.dry COP =ma =po =wei =CAUS =CONS =HAB 'Even after it rained, (the leaves) stay dry because a temporary house certainly exists.’ Example (99) shows the conservative with the verb of motion, hayi 'to pass'. (99) kaye 1SG hiyehè here Boa Vista Boa Vista =hè =LOC cinco five semana week yeh= 1DL.SUBJ= apè= UNSP= hayi =ma =po =le pass =CAUS =CONS =PRS 'I (am) here in Boa Vista. We (dual) are spending five weeks.’ 5.4.4 Perfective =a As mentioned in §5.1, the PLURACTIONAL =a has a homophonous morpheme that occurs in a different slot after the adverbials, c.f. example (55), here repeated as (100). It was observed that this morpheme occurs in complementary distribution with the other aspects listed in this section and it will be labeled as PERFECTIVE, the same way as Ferreira (2017). (100) ya= pulu =haiki =a 1SG= lie.in.hammock =certainly =PFV ‘I’m certainly lying in the hammock’ (CS02b) (101) ya= pulu =a 1SG= be.in.hammock =PFV ‘I’m in the hammock.’ (102) ĩe apè= helexi =a =le blood UNSP= foamy =PFV =PRS ‘The blood is thin.’ 56 5.5 Middle voice =o In addition to the morphemes of valency change explained in §5.1, Ninam of Upper Mucajaí also presents the morpheme =o in its constructions. The first hypothesis is to consider this morpheme as a middle voice, the same way as other linguists such as Ramirez (1994) and Lizot (2004: 279) have described it. In the story of the spider monkey, it is told that the monkey grabbed a nut. Specifically, in example (103), it is mentioned that after the monkey had hit itself while hitting the nut, the nut opened. The whole context can lead to the use of a middle voice in the construction. (103) ha= SEQ= yatiki hit =la =PLU =mo =REFL =yo =REC =n =after kahiki mouth ha= HEAR= lelehe =o =elei open =MID.VOI =REM.PST 'After hitting himself and it (the Brazil nut), the mouth (of the nut) opened.' Another analysis for this morpheme is the one proposed by Migliazza (1972) who argues that the reflexive =mo is composed of the causative =ma and what he labeled as [+reflexive] =o. This would mean that the reflexive =mo is constructed from the notion of the causative where the valency is increased, thus filling the argument space that was added by using the reflexive. It would also explain why the reflexive =mo and the middle voice =o do not co-occur. However, examples such as (74) and (104) demonstrate that the causative can cooccur with reflexive =mo and the middle voice =o, respectively. Thus, reducing the probability of coalescence. Therefore, there is another reason why =mo and =o do not cooccur. Most probably because both are valency reducing mechanisms that work in different ways; with the reflexive, the subject and the object are clearly the same person; while with the middle voice, one of the arguments is not mentioned. Example (104) has the verb luku 'to enter a hole'. An intransitive verb that carries the idea of a movement. The verb is followed by the causative =ma which increases the valency of the verb, so that two participants are expected. Then, the middle voice =o decreases the valency, leaving the 3rd plural pronoun pèk= as the only argument. 57 (104) thootho =hèm pèk= apè= lũku =ma =o =wei liana =DIR 3PL.SUBJ= UNSP= enter.in.a.hole =CAUS =MID.VOI =HAB ‘They force themselves into in (the water), towards the liana.’ In addition to the middle voice interpretation, the morpheme =o can be interpreted as 'to place oneself in position' or 'to be in a position' (Lizot 2004: 279). This interpretation could fit single-argument constructions, specifically with stative verbs. The function of the morpheme =o is seen when compared to other morphemes occurring in the same slot. In example (85), here repeated as (105a) the RESULTATIVE =layo implies a change of state, e.g., the person was sitting and stood up. In example (105b) the =o brings the idea that the subject was in that position. (105) a. ya= pè= ila 1SG= UNSP= stand ‘I stood (from being sit).’ =layo =RES =he =REC.PST2 b. ya= pè= ila 1SG= UNSP= stand ‘I stood up myself.’ =o =MID.VOI =he =REC.PST2 Based on these data, the label initially proposed is MIDDLE VOICE when the morpheme =o occurs with transitive verbs as it contributes the meaning that the person placed him/herself in the situation being reported. 5.6 Verbal directionals and locatives According to the data, Ninam of Upper Mucajaí also has morphemes that encode direction and location. These morphemes occur in the same slot, after the aspect slot, and before the tense. 5.6.1 Verbal directionals In Ninam, verbal directionals correspond to a separate category from adverbials due to the difference that arises when comparing the position of the evidentiality marker within the verbal complex in the presence of any adverbial versus a verbal directional. For the adverbials, the evidentiality marker changes its position, c.f. §4.3.5., whereas no such alteration occurs when directionals are present. Therefore, they must be classified as an 58 independent grammatical category. The directionals found in the Ninam data are =ima and =heli ITIVE. VENITIVE These morphemes indicate the direction in which the action is performed and although they tend to occur with verbs of motion such as ha ~ hu ‘to move', kope 'to return', and lele 'to run', they are not obligatory. This is observed by comparing examples (106) and (107) which have the same verb, hu 'to move'. In the first example, the direction of the movement is not stated. In contrast, in the second example, the verb is followed by =ima VENITIVE, which indicates the direction towards a reference. Note that verbal directionals must also be differentiated from the postpositions occurring in the noun phrase, namely =elihè and =hèm DIRECTION. (106) ya= hu =ithe 1SG= move =FUT ‘I will go.’ (107) naa =ye =elihè ya= ũku pè= hu =ima =he mother = 1SG. POSS =DIR 1SG= cry UNSP= move =VEN =REC.PST2 ‘I came to cry for my mother.’ The verbal directional =heli ITIVE, observed in examples (108) and (109), indicates that the movement is outward, away from the deictic reference, which is the subject of the sentence. In (108), the narrator explains that he will return home, and the reference is a direction away from the city where he is at the time of the narration. This is consistent with (109), where a Ninam ran away after seeing the dead spider monkey. (108) ilihi =hèm yehek= kope =heli =ithe forest =DIR 1DL.SUBJ= return =ITIV =FUT ‘We (dual) will return to home (lit. to the forest).’ (109) ha= taa =lu =n ha= kõ= le =heli =lei SEQ= see =RES =after HEAR= again= run =ITIV =REM.PST ‘After seeing it, he ran away.’ 5.6.2 Verbal locatives In addition to directionals, Ninam presents morphemes coding spatial features, that is, verbal locative morphemes that encode the idea of where the narrated action takes place. The verbal locatives in Ninam can be a sequence of two morphemes, a morpheme indicating the plane 59 and another morpheme indicating the level, c.f. Table 20. Note that verbal locatives must also be differentiated from the postposition occurring in the noun phrase, namely =hè LOCATIVE. Table 20 – Morpheme breaks in the locatives of Ninam of Upper Mucajaí PLANE =la Horizontal =po Vertical (homophone of CONSERVATIVE) LEVEL =li =kili~kli =ti Up (homophone of RESULTATIVE) Down With an obstruction in between that hinders the line of sight (near) Considering the terms above, the morphemes indicating plane occur in the same position in which aspect markers occur. This means that the morphemes indicating the plane are in complementary distribution with the aspect markers. Following the plane morphemes, the morphemes indicating level occur. The sequence of the morphemes of plane and level allow concepts such as ‘upriver’ and ‘downriver’ to be constructed, c.f. Table 21. Table 21 – Locatives of the horizontal plane in Ninam of Upper Mucajaí =la =li Distant Upriver =la =ti =la =kili ~ =kli =la =lo Not so far On the other side of the river Downriver Around These morphemes are used to refer to the distance from the speaker's perspective. Figure 8 presents the actual space where the Upper Mucajaí village is located. A Ninam speaker will use the sequence =la=li to refer to upriver distant locations, e.g. distance from the Upper Mucajaí village to the Pewa u village – red line. The speaker will use the sequence =la =ti if the location is not far, e.g. distance from the Upper Mucajaí village to the Polapiiu village – yellow. If it is a near location, the speaker will use the sequence =la=lo. 60 Figure 8 – Map showing the villages around the Upper Mucajaí village Following the geography known to the Ninam, two sequences of morphemes refer to the flow of the river. The sequence =la=li refers to ‘upriver’ and =la =kili ~ =kli refers to ‘downriver.’ Examples (110) to (113) show the first set of morphemes discussed in the previous paragraphs. (110) Kõmoli themu =hè yamak= lam apè= hu =la =li all day =LOC 1PL.SUBJ= hunt UNSP= move =horizontal =up =wei =HAB ‘Every day we go hunting far upriver.’ (111) Kõmoli themu =hè yamak= lam apè= hu =kli =wei all day =LOC 1PL.SUBJ= hunt UNSP= move =down =HAB ‘Everyday we go hunting downriver.’ (112) Kõmoli themu =hè yamak= lam apè= hu =la all day =LOC 1PL.SUBJ= hunt UNSP= move =horizontal =ti =wei =near =HAB ‘Every day we go hunting on the other side of the river.’ 61 (113) Kõmoli all themu =hè yamak= lam apè= hu =la day =LOC 1PL.SUBJ= hunt UNSP= move =horizontal =lo =wei =around =HAB ‘Every day we go hunting around.’ The frequency of locatives in the data collected was scarce. This is presumably explained because the data were collected in the city of Boa Vista and not in the Upper Mucajaí village so spatial references such as upriver and downriver are not coherent with the location of collection. Example (114) is the answer to a question like hapeli hèm kama poaka kua? 'Where is his axe?' In the answer, the speaker indicates the specific location by mentioning the house followed by the postposition of direction =hèm and then, in the verbal complex, the locative =la=ti indicates that it is not that far. This contrasts with examples in (115) and (116), which only present the morpheme =kli indicating downriver as the location. (114) yano =hèm ku =la =ti house =DIR COP =horizontal =near ‘It is not so far, in the direction of the house.’ (115) yalu =mo =kli =xo wash =REFL =down =COMIT ‘Let’s go bathe downriver.’ (116) ya= pè= kõ =kli =he 1SG= UNSP= return =down =REC.PST2 ‘I returned from being downriver.’ The other possible sequence is that of the vertical plane with the level morphemes, as presented in Table 22. These morphemes that convey the idea of things that are above or below. Table 22 – Locatives of the vertical plane in Ninam of Upper Mucajaí =pe Below =po =ti Below Hidden =po =li Above 62 By using the morpheme =pe, the speaker indicates that the subject of the sentence is located in a sunken place or below the reference point; therefore, it has been labeled as below. This can be used with places like the river or hollows as presented in the data (117) and (118) respectively. (117) Mau =hèm ipa kamixa apè= ke =pe =he river =DIR 1SG.POSS clothes UNSP= fall =below =REC.PST2 ‘My clothes fell in the river.’ (118) paka =hèm hesi ya= pè= ke =ma =pe =he hole =DIR glass 1SG= UNSP= fall =CAUS =below =REC.PST2 ‘I threw the glass in a hole.’ Using of the morpheme =po =ti, the speaker also means that the subject of the sentence is inside something like a hole, a well, or at home. However, different from =pe presented in the previous paragraph, the morpheme =po =ti adds that the subject cannot be seen. In examples (119), the locative follows the verb and is labeled as 'inside'. (119) aho liplosi =hèm ku =po =ti 2SG.POSS book =DIR COP =CONS =near ‘It is inside your book.’ Examples in (120) describe the upward and downward movement that occurs when breathing. Example (120a) refers to the movement of the chest. Then, in (120b), the speaker refers to the movement within a person's chest because of breathing. Since it is understood that the movement is produced by something internal that is not seen, the speaker used the locative =po =ti for this purpose. (120) a. paluk apè= yutyut =mo =le chest UNSP= move.upward.downward =REFL =PRS ‘The chest goes up and down.’ (CS04) b. apè= yutyut =mo =po =ti UNSP= move.upward.downward =REFL =CONS =near ‘It goes up and down in the inside.’ (CS04) 63 The morpheme =po =li indicates that something is located above. In the Ninam spatial context, there are not a lot of options for things that are high or above – the sky, the roof of a house, or branches on a tree. Example (121) shows the morphemes =po=li locating a bird that is up in a tree. Example (122) refers to something that was hunted and now hangs in a tree. (121) wii =tihi poko =hèm yãpek =hèm kĩli ku =po =li tree =CLN:tree branch =LOC nest =LOC bird COP =CONS =up ‘The bird is in the nest up on the branch in the tree.’ (122) apè= yete =po =li UNSP= be.joined =CONS =up ‘It is hung up above.’ 5.7 Tense, aspect, mood, infinitive and comitative markings Albright’s (1965) initial analysis suggests that Ninam of Upper Mucajaí has aspect markings instead of tense markings. However, the data analyzed does not suggest this label as the only one for morphemes occurring in this slot. This section describes the tense, aspect, mood, and infinitive morphemes found in complementary distribution on the right side of the verbal complex in main sentences. 5.7.1 Tense markers Ninam presents different markers that encode tense information. The tense system is shown in Table 23 and further described in the subsequent subsections. Table 23 – Past tense continuum in Ninam of Upper Mucajaí PAST PRESENT Remote Recent Pre hodiernal Hodiernal =elei =ma/=he =lexi =xi FUTURE Distant =le/Ø =ithè 5.7.1.1 Remote past Considering the system as a continuum, the description begins with the morpheme =elei ~ =lei that was found in mythical narratives as in examples (123) and (124), and in other narratives as in (125). It has been labeled as REMOTE PAST because it refers to events far from the present. 64 (123) ha= huxmili =layo =lei HEAR= be.dead =RES =REM.PST ‘It was dead’ or ‘it died.’ (124) pèk= ha= laiwa =elei 3PL.SUBJ= HEAR= dance =REM.PST 'They danced.’ (125) Teus God thã word =elihè ya= kẽa =wei sĩuthai ta= ku =LOC 1SG= work =INF a.little.bit VIS= COP =o =lei =MID.VOI =REM.PST ‘I worked a little bit on the Word of God.’ 5.7.1.2 Recent past Following the continuum presented above, two other markings for the past were found in the data, =ma and =he. Both morphemes are employed in narratives, and their use is related to the use of evidentials. All cases realized with the morpheme =ma do not require the use of evidentials. In contrast, all cases realized with the morpheme =he require the use of evidentials. This same behavior was observed in Xamatauteri (Ramirez 1994: 311–312) and Yanomama (Ferreira 2017: 306). In this work, they are labeled as RECENT PAST 1 and RECENT PAST 2 to differentiate them from the REMOTE PAST. In relation to the aspects, the recent past =ma only occurs with the resultative aspect, as shown in examples (87), here (126), and (127). (126) ipa ilihi 1SG.POSS forest =hè sete ano ya= kẽa =hela =LOC seven year 1SG= work =ITIV =yo =ma =RES =REC.PST1 ‘I worked back home (lit. in my forest) for seven years.' (127) ya= naka =le =ma =he 1SG= call =RES =REC.PST1 =3PL.SUBJ ‘They called me.’ On the other hand, the morpheme =he can co-occur with the three aspects =layo, =le, and =ki, as shown in example (83), numbered here as (128), (93), numbered here as (129), 65 and (130). The morpheme =he was labeled by Tracy (1966), Albright (1970), Swain (1971), and Borgman (1999) as a COMPLETED ASPECT. Since this morpheme does not occur in present or future readings, the label that seems most consistent with its occurrences is PAST. As for the morpheme =ma, languages such as Xamatauteri (Ramirez 1994), Yãnomãmɨ (Lizot 2004), and Yanomama (Ferreira 2017) present the same morpheme, but it has only been identified as a past tense without further details. (128) Xama apè= lothotho =layo =he tapir UNSP= be.strong =RES =REC.PST2 ‘The tapir got strong.’ (CS02) (129) sĩuthai apè= ta =le =ha =hei ai =pèku =n a.little.bit UNSP= see =RES =REC.PST2 =3PL.SUBJ other =PL =ERG ‘Others saw a little bit.’ (130) hoithehe cinquenta porcento now fifty percent =elihè vídeo yeh= apè= tha la =LOC video 1DL.SUBJ= UNSP= birth =ma =ki =he =CAUS =INCEP =REC.PST2 ‘Now we (dual) have brought to existence 50% of the video.' 5.7.1.3 Prehodiernal and Hodiernal Two more past tense markings were identified by Swain (1971: 16). The morpheme=lexi is labeled here as PREHODIERNAL and the morpheme=xi is labeled here as HODIERNAL past. According to the descriptions provided by Swain (2002), both tense markings are only used with the evidential ta= VISUAL, as seen in examples (131) and (132). (131) Aikam =theli ta= huxmi =layo =lexi Aikam =person VIS= be.dead =RES =PRE.HOD ‘The Aikamtheli died (before today).’ (CS04) (132) Mene e ta= hoke =yo =xi mother 3SG.POSS VIS= travel =RES =HOD ‘His mother traveled (today).’ (CS02b) 66 5.7.1.4 Present tense Ninam indicates a reading in the present tense through zero marking, as in example (133), but the present tense is also encoded by the morpheme =le, labeled as PRESENT. The morpheme =le carries the idea of present tense and progressive aspect, c.f. (134) and (135). (133) apè= ohi UNSP= be.hungry ‘He is hungry.’ (SA65) (134) apè= ohi UNSP= be.hungry ‘He is (currently) hungry.’ (SA65) (135) hiyehè gravação =elihè ya= pè= kẽa =le here recording =LOC 1SG= UNSP= work =PRS ‘I am working on the recording.' =le =PRS Note that the present tense marker =le does not co-occur with the RESULTATIVE nor the INCEPTIVE aspect markers. Nevertheless, it can occur with the CONSERVATIVE =po, which gives the idea that the action is continued, c.f. example (65), numbered here as (136). (136) naa =ho =n yalaka apè= wa =po =le mother =2SG.POSS =ERG fish UNSP= eat =CONS =PRS ‘Your mother keeps eating fish.’ 5.7.1.5 Distant future In Ninam of the Upper Mucajaí, the future tense is marked through the morpheme =ithe labeled FUTURE. This morpheme only occurred when the informant was asked what he would do when he returned home after his stay in Boa Vista. Examples (137) and (138) have the copula followed by the morpheme =o coding middle voice and the future tense marker =ithe. (137) yãno house =hèm =DIR ya= 1SG= walo =yapa arrive =back =ke =thehe =INCEP =when haximi watu =wei patapata ku =o =ithe rest =NMLZ big~AUG COP =MID.VOI =FUT ‘When I get back home, there will be a huge rest.' 67 (138) ipa ulu =ye =pèk thẽlema =wei ki =o =ithe 1SG.POSS child =1SG =PL take.care =NMLZ COP =MID.VOI =FUT ‘I will take care of my children.’ Although the morpheme indicating the future was mostly found accompanying the COPULA, this morpheme can also accompany stative and dynamic verbs. The morpheme =ithe occurs with descriptive verbs such as xiho ‘to be happy’, in (139), and positional verbs such as hĩu ‘to be in a near angle’ in (140). It also occurs with dynamic verbs as the intransitive construction in (108), here (141), and the transitive constructions in (142) and (143). (139) naxi =kok wa= xuma =uhè, ya= xiho cassava =CLN:root 2SG= send =COND 1SG= be.happy =yo =ithe =RES =FUT ‘If you send the cassava, I will be happy.’ (140) apièm hũu =ke =heli =ithe airplane be.in. near angle =INCEP =ITIV =FUT ‘The airplane will take off.’ (141) ilihi =hèm yehek= kope =heli =ithe forest =DIR 1DL.SUBJ= return =ITIV =FUT ‘We (dual) will return to home (lit. to the forest).’ (142) xama wa= hete =uhè, wa= ta =le =ithe tapir 2SG= hunt =COND 2SG = see =RES =FUT ‘If you hunt a tapir, you will see it.’ (143) Xakarina =elihè ya= thã= wãli =ithe Jacqueline =DIR 1SG= word ask =FUT ‘I will ask Jacqueline about the word.’ 5.7.2 Habitual aspect In the procedural texts associated with Yanomami culture, such as making cassava drink, c.f. example (144), or fishing when they have a party, c.f. example (145), the morpheme =wei can be read as ‘generally done’. Although the information it encodes is clearly aspectual, the morpheme =wei behaves like a tense marker, because of its position at the end of the morpheme chain, and, in addition, it hosts the clitic =he 3rd plural subject, a quality exclusive to morphemes that indicate tense, see §5.7.6. 68 (144) thuwe =un nax =kok pè= uka =wei woman =ERG wild.yucca =CLN:root UNSP= pluck = HAB ‘The woman plucks wild yucca.’ (145) yalami yalaka =pèk apè= noma =wei many fish =PL.AN UNSP= be.dizzy =HAB ‘Many fish are dizzy.’ 5.7.3 Determined marker According to Swain (2002), the morpheme =paken indicates a 'potential future' or that the subject is determined to perform the action described. Swain (2002) also mentions that this morpheme is only used with 1st person and does not occur with the ‘dynamic focus,’ called RESULTATIVE aspect in this work. It also does not co-occur with the adverbial of negation. Examples (146) to (147) show the morpheme =paken with stative and dynamic verbs. (146) hena =thehe yehek= pulu =o =paken tomorrow =when 1DL.SUBJ= lie.in.hammock =MID.VOI =DETERM ‘We (dual) are determined to lie in the hammock.’ (147) ya= hoke =paken 1SG= travel =DETERM ‘I am determined to travel.’ 5.7.4 Infinitive maker According to the data, the infinitive is indicated by the morpheme =wei, previously labeled as habitual aspect, c.f. 5.7.2. This morpheme is attached directly to the verb stem without any other morphology as in example (148a) and (148b). (148) a. uka pluck =wei =INF ‘to pluck’ b. noma be.dizzy =wei =INF ‘to be dizzy’ It is also used in constructions with two verbs. In these cases, other morphology can be attached to it. In example (125), here (149), the verb stem kẽa ‘work’ also hosts the pronominal reference for 1st singular ya=. 69 (149) Teus God thã word =elihè ya= kẽa =wei sĩuthai ta= ku =LOC 1SG= work =INF a.little.bit VIS= COP =o =lei =MID.VOI =REM.PST ‘I worked a little bit on the Word of God.’ 5.7.5 Comitative =xo The morpheme =xo has been labeled as INCLUSIVE in Xamatauteri (Ramirez 1994: 204-207) and as ADDITIVE in Yanomama (Ferreira 2017: 403). It has been described as a morpheme that occurs in the noun phrase and has the function of coordinating, so it can be translated as 'and', 'with' and 'in the company of.' In Ninam, when the morpheme =xo accompanies noun phrases, it coordinates them; however, when it accompanies the verb phrase it gives the idea of an invitation to do something together. (150) ai hutu =sik tha =kõ =o =xo other farm =CLN:planted do =again =MID.VOI =COMIT 'Let’s make another farm again.' (151) hemei ipa hutu tha =a =xo let's.go 1SG.POSS farm do =PFV =COMIT 'Let's go, make my farm.' 5.7.6 The morpheme =i According to the data, the morpheme =i is found in complementary distribution with the morphemes marking tense. However, its meaning is still unclear. For other Yanomami languages, a similar morpheme has been labeled as 'non-aspectual' (Borgman 1990), 'imperfective' (Goodwin 1990: 98), evidential ‘witnessed' (Ramirez 1994: 316), 'epistemic modality' and 'modality of insistence' (Lizot 2004: 123), 'dynamic' (Ferreira 2011, 2017: 191). 5.7.6.1 Tense marker Swain's (1971) study indicates that the morpheme =i can be read as 'to be,' 'about to be,' and 'potential.' Meanwhile, Albright's (1970) study indicates that the morpheme =i has a homophone. According to her study, one of them would indicate an action in the recent past, while the other would indicate the present tense in process. Albright (1970) also explains that 70 in conjunction with the evidentials ta= VIS and ha= HEAR the reading should be in the past tense, c.f. (152) and (153). On the other hand, if it occurs with the evidential apè= ~ pè= UNSP, the reading should be in the present tense, c.f. example (154). If there are no evidentials, the reading can be in the present tense or the immediate future, c.f. example (155). In addition to this, constructions with past tense reading will be the only ones that will host evidentials. Following that approach, at least two labels would be necessary, one to indicate reading in the past and one for reading in progress. (152) Kahowa pemi =o =thehe 2SG non.existence =MID.VOI =when popa =ho ta= father =2SG.POSS VIS= nini =yo =i be.hurt =RES =PST ‘While you were away, your father was hurting.’ (SA70) (153) Antonio ha= waro =ki =i Antonio HEAR= arrive =RES =PST ‘Antonio arrived.’ (SA70) (154) Them apè= wakixi =i day UNSP= be.cloudy =?? ‘The day is cloudy.’ (SA70) (155) Naa =ho =elihè ya= uku =i mom =2SG.POSS =DIR 1SG= cry =IMM ‘I’m about to cry because of your mother.’ (SA70) 5.7.6.2 Imperative reading According to Swain (2004), the morpheme =i can also have an imperative reading when the construction has the 2nd person. Apart from the use of the 2nd person, no other morphological distinction is made to differentiate between the imperative mood and the tense marking function. (156) wa= ku =yo =i 2SG= speak =RES =IMP ‘Speak!’ 71 (157) wa= ila =layo =i 2SG= stand =RES =IMP ‘Stand!’ (CS02b) (158) wamak= ũku =yo =i 2PL= cry =RES =IMP ‘Cry!’ 5.8 Pronominal reference to the ergative 3rd dual and plural As mentioned in §4.1.1.3, the pronominal reference indicating the 3rd dual and plural as the subject of the transitive sentence is placed on the right side of the verbal complex. From the data collected, it is observed that the morpheme =he(i) 3rd plural occurs near the tense markings.4 It occurs infixed into =elei REMOTE PAST marker and =wei HABITUAL marker, as shown in examples (159) and (160), before the future tense marker, as shown in example (161), or after the recent past tense marker, as shown in examples (162) and (163). (159) pata =pèk =un wake hipo =ha =lã =elai old =PL.AN =ERG fire own =HEAR =NEG =REM.PST:3PL.SUBJ ‘The elders had no fire.’ (160) Ũhu dois pèk =un apè= tha =wei DEM two 3PL =ERG UNSP= do =HAB:3PL.SUBJ ‘These two make it.’ (161) Wãlo =pèk =un nakamthok pexima =he =ithe man =PL =ERG hammock want =3PL.SUBJ =FUT ‘The men will want the hammock.’ (CS02) (162) missionario =pèk =un ya= thẽlema =ma =he missionary =3PL =ERG 1SG.OBJ= help =REC.PST1 =3PL.SUBJ ‘The missionaries helped me.’ (163) apè= ta =le =ha =hei UNSP= see =RES =REC.PST2 =3PL.SUBJ ‘They saw it.’ 4 Accordding to Swain (1971: 14), the marker for 3rd dual subject =pu also occurs in this position. However, no examples were found. 72 6. Conclusions This research aimed to provide an initial description of the verbal morphology of Upper Mucajaí Ninam, with a specific focus on the morphology of the verbal complex in the main sentence. This aim was achieved through the collection, transcription, and analysis of Ninam language data, in parallel with the review of related literature. This enabled me to observe common features with other Yanomami languages, as well as the singularities of the Ninam language itself. Throughout this study, five preverbal slots and at least nine postverbal slots were identified, revealing a complex morphosyntactic system that comprises indexation of subject and object, evidentials, adverbials, valency change, aspect marking, directionals, middle voice, and tense marking. On the left side of the verb stem, the most noteworthy findings are the subject and object indexing system, as well as the system of evidentials. On the right side, the set of locatives and tense markings showed a variety of morphemes that enriched the possibilities for sentence formation in Ninam. Additionally, a significant similarity with other languages within the Yanomami family was observed, thereby enriching the understanding of the data. In terms of research contribution, this work is considered crucial for ongoing linguistic analysis within the Yanomami family. Specifically for Ninam, this study addresses an existing gap, as there were no specific studies on morphosyntax, unlike other levels of analysis. Also, the adopted framework-free approach (Haspelmath 2008) allowed the study of the language on its terms, revealing singularities as mentioned earlier. Looking ahead to future research, various areas are suggested. Building on the proposals of this work, exploring the verb in subordinate and relative clauses is recommended. Furthermore, based on the acquired knowledge about verb morphology, the study of other grammatical classes to describe their structure is proposed. There is also an emphasis on the need for deeper research into the evidential system to identify possible variations. Additionally, a detailed study of tense marking is suggested, as some morphemes, such as =i and =wei, appear to have more complex functions than can be deduced from current data. 73 This study not only provides a detailed description of the verb in the main sentence in Ninam del Alto Mucajaí but also lays the groundwork for future research that will expand our understanding of morphosyntax and other linguistic aspects within the Yanomami family. 74 References Aikhenvald, Alexandra & R.M.W. Dixon. 1998. Evidentials and areal typology: a case-study from Amazonia. Language Sciences 20. 241-257. Aikhenvald, Alexandra & Robert Dixon. 1999. Other small families and isolates. In Alexandra Aikhenvald & Robert Dixon (eds.), The Amazonian languages, 341-384. Cambridge: Cambridge University Press. Aikhenvald, Alexandra. 2003. Evidentiality in typological perspective. In R.M.W. Dixon & Alexandra Y. Aikhenvald (eds.), Studies in Evidentiality, 1-32. Amsterdam: John Benjamins. Albright, Sue R. 1965. Aykamteli higher-level phonology. Anthropological Linguistics (7). 16–22. Albright, Sue R. 1970. Kind of knowledge, information source, location and time in Siriana predicates. Seminar Papers Series (1). 1–19. Anderson, L. B. 1986. Evidentials, paths of change, and mental maps: typologically regular asymmetries. In W. Chafe & J. Nichols (eds.), Evidentiality: the linguistic coding of Epistemology, 273–312. Norwood: Ablex. Barker, Jamers. 1963. Diccionario Guaika, and other mimeographed grammatical notes. (Manuscripts). New Tribes, Venezuela. Barker, James. 1979. Una gramática técnica de la lengua Shamatari (notas por Estaban Emilio Mosonyi). Boletín Indigenista Venezolano: Tomo XVIII, XVIII (15): 193-226. Borgman, Donald. 1959a. Language data preliminary conclusions. Unevengelized Field Mission (UFM). Non-published. Borgman, Donald. 1959b. Morphological structure of the Shirishaná dialect. Unevengelized Field Mission (UFM). Non-published. Borgman, Donald. 1990. Sanuma. In Desmond Derbyshire & Geoffrey Pullum (eds.), Handbook of Amazonian languages, vol 2, 17-231. Berlin: Mouton de Gruyter. Borgman, Donald, Sandra Cue, Sue Albright, Merril Seeley & Joseph E. Grimes. 1965. The Waican languages, Anthropological Linguistics 7 (7). 1-4. Campbell, Lyle. 1997. American Indian languages – the historical linguistics of Native America. New York: Oxford University Press. Derbyshire, Desmond. 1987. Morphosyntactic areal characteristics of Amazonian languages, International Journal of American Linguistics 53 (3). 311–326. 75 Derbyshire, Desmond & Doris Payne. 1990. Noun classification systems of the Amazonian languages. In Doris Payne (ed.), Amazonian Linguistics - Studies in Lowland South American Languages, 243–271. Austin: University of Texas Press. Dworecka Autuori, Joana. 2013. Aspectos da Fonologia da Língua Sanumá (Yanomami). Boa Vista, RR: Universidade Federal de Roraima at Boa Vista MA thesis. Early, John D. & Peters, John F. 2000. The Xilixana Yanomami of the Amazon: History, social structure, and population dynamics. Gainesville: University Press of Florida. Eberhard, David M., Gary F. Simons, & Charles D. Fennig (eds.). 2022. Ethnologue: Languages of the World. Twenty-fifth edition. Dallas, Texas: SIL International. Online version: http://www.ethnologue.com. Ferreira, Helder, Ana Machado & Estêvão Benfica. 2019. As línguas Yanomami no Brasil: diversidade e vitalidade. São Paulo: ISA - Instituto Socioambiental. Ferreira, Helder. 2011. Esboço gramatical do Yaroamë: Língua Yanomami falada na Serra do Pacu/RR. Rio de Janeiro: Museu do Índio/Prodoclin. Ferreira, Helder. 2017. Yanomama clause structure - part I. Utrecht: LOT. Gómez Goodwin, Gale. 1990. The Shiriana dialect of Yanam (Northern Brazil). Ph.D. thesis – Columbia University. Gómez Goodwin, Gale. 2018. Nonverbal predication in Ninam (northern Brazil). In Spike Gildea & Fernando Zúñiga, Typological Studies in Language, 245 – 262. Amsterdam / Philadelphia: John Benjamins Publishing Company Governo Federal do Brasil. 2023. Relatório Missão Yanomami. https://www.gov.br/saude/ptbr/assuntos/noticias/2023/fevereiro/arquivos/RelatorioYanomamiversao_FINAL_07_02. pdf (05 May 2023). Haspelmath, Martin. 2008. Framework-free grammatical theory. https://www.eva.mpg.de/fileadmin/content_files/staff/haspelmt/pdf/Frameworkfree.pdf Infante Peña, Victoria Elizabeth. 2023. Evidentiality in Ninam of Alto Mucajaí. LIAMES: Línguas Indígenas Americanas. V.23. https://periodicos.sbu.unicamp.br/ojs/index.php/liames/article/view/8673977/32578 Instituto Nacional de Estadísticas. 2011. Censo Nacional de Población y Vivienda 2011 Empadronamiento de la Población Indígena. http://www.ine.gob.ve/documentos/Demografia/Censo2011/pdf/EmpadronamientoIndig ena.pdf (03 May 2023). Koch-Grunberg, Theodor. 1923. Vom Roraima zum Orinoco vol 4. Berlin: Dietrich Reimer. 76 Lizot, Jacques. 1996. Introducción a la lengua yanomamɨ: Morfología. Caracas: Vicariato Apostólico de Puerto Ayacucho/UNICEF. Lizot, Jacques. 2004. Diccionario enciclopédico de la lengua Yanomami. Puerto Ayacucho: Vicariato Apostólico de Puerto Ayacucho. Migliazza, Ernest & Joseph Grimes. 1961. Shiriana Phonology. Anthropological Linguistics 3(6). 31–41. Migliazza, Ernest. 1972. Yanomama grammar and intelligibility. Ph.D. thesis ‒ Indiana University, Bloomington. Migliazza, Ernest. 1985. Languages of the Orinoco-Amazon region: current status. In Harriet Manelis & Louisa Stark (eds.), South American Indian Languages – retrospect and prospect, 17-139. Austin: University of Texas Press. Missão Evangélica Caiuá. 2023. Relatorio de Visita ao Distrito Sanitário Indigena Yanomami. https://jornalggn.com.br/wpcontent/uploads/2023/02/RELATORIO_VISITA_AO_YA_assinado.pdf (05 May 2023). Mosonyi, Esteban Emilio. 2014. El misionero James Barker: ¿El pánini de la lengua yanomami?. Revista Brasileira de Linguística Antropológica (6)2. 543‒559. Povos Indígenas no Brasil. 2023. Yanomami. https://pib.socioambiental.org/pt/Povo:Yanomami (03 May 2023). Ramirez, Henri. 1994. Le Parler Yanomamɨ des Xamatauteri. Aix-En-Provence: Université de Provence thesis. Swain, Carole. 1971. A sketch of the grammar of the Shilishana dialect of Shiriana. SIL. Non-published. Swain, Carole. 2002. Apostila do Curso para a Aprendizagem da Língua Ninam. MEVA. Non-published. Swain, Carole. 2002b. Curso para a Aprendizagem da Língua Ninam – Lista de dados. MEVA. Non-published. Swain, Carole. 2004. Dictionário Ninam - Português. MEVA. Non-published. Tinoco, Rodrigo. 2023. Mapas da Terra Yanomami. Non-published. Tracy Votey, Frances. 1966. The Phonology and Outline Grammar of the Aikamtheli Dialect of Shiriana, with notes of the Other Dialects. Non-published MA thesis, Philadelphia: University of Pennsylvania. 77 Appendices 6.1 Appendix 1: The story of the spider monkey (mythical narrative) (1) Yalimi =n hoali =mok ha= xe =a =elei spider.monkey =ERG Brazil.nut =CLN:seed HEAR= hit =PLU =REM.PST 'The spider monkey hit the Brazil nut' (2) Wii =tihi poko hewe =hè ha= xe =a =elei tree =CLN:wood branch dry =LOC HEAR= hit =PLU =REM.PST ‘He hit it on the dry branch.’ (3) Ũhu DEM utha =k ha= yatiki =la =mo =yo finger =PL HEAR= hit.with.wood.stick =DIST =REFLEX =RES =lei =REM.PST ‘He hit (its own) hand with a wood stick.’ (4) Ha= HEAR= yatiki hit.with.wood.stick =la =DIST =mo =REFLEX =yo =RES =n =after kahiki ha= lelehe =o =elei mouth HEAR= yawn =MID.VOI =REM.PST ‘After he hit (his own) hand with a wood stick, the nut opened (lit. the mouth opened).’ (5) Yalimi =n hoali =mok ha= ke =ma =elei spider.monkey =ERG Brazil.nut =CLN:seed HEAR= fall =CAUS =REM.PST ‘The spider monkey threw the Brazil nut.’ (6) Ha= HEAR= ke fall =ma =CAUS =lu =RES =n =after itho going.down =ha =SEQ =kõ =again =ma =CAUS =ku =INCEP =n =after mok= CLN:seed= te grab yapa back =ha =SEQ =kõ =again =li =RES =n =after xe =ha =kõ =elei hit =HEAR =again =REM.PST 'After dropping it, after going down again, after taking back the nut, it hit it again' 78 (7) Xe hit =ha =SEQ =kõ =again =ke =RES =n =after yalimi spider.monkey pihi =kãi ha= yatiki =layo =lei mind =SIM HEAR= hit.with.wood.stick =RES =REM.PST 'After hitting it again, the spider monkey was hit with a wood stick simultaneously.' (8) Yalimi =n pihi =kãi ha= ke =yo =lei spider.monkey =ERG mind =SIM HEAR= fall =RES =REM.PST ‘The spider monkey fell simultaneously.' (9) Ha= ke =yo =n ha= SEQ= fall =RES =after HEAR= 'After it fell, it hit with a wood stick.' (10) Ha= HEAR= yatiki hit.with.wood.stick huxmili =layo =lei =be.dead =RES =REM.PST 'It died.' (11) Ninamu =n yalimi huxmili ha= ta =elei person =ERG spider.monkey dead HEAR= see =REM.PST 'A person saw the dead spider monkey.' (12) Ha= ta =lu =n ha= kõ= le =heli =lei SEQ= see =RES =after HEAR= again= run =ITIV =REM.PST 'After seeing it, he ran away' =ke =lei =INCEP =REM.PST 79 6.2 Appendix 2: The word (narrative) (1) ku =pè =haiki =le =UNSP =certainly =PRS ‘ok (lit. certainly it is) COP (2) yutuhè long.time.ago ya= 1SG= oxe young.man =a =PFV =thehe =when liplosi book ya= 1SG= tũlu be.written tapo know =maha =CONCS =lã =NEG =noi =?? missionário missionary =pèku =3PL =n =ERG ya= 1SG.OBJ= miyawe be.smart =ma =CAUS =le =RES =ma =REC.PST1 =he =3PL.SUBJ 'A long time ago, when I was a boy, even though I couldn't write, the missionaries made me smart. ' (3) inaha ku =yalo hiyehè gravação =elihè ya= pè= kẽa =le thus COP =EXPLV here recording =DIR 1SG= UNSP= work =PRS 'Thus, I am working on the recording,' (4) yutuhè professora Xakalina ya= ta =le long.time.ago female.teacher Jacqueline 1SG.SUBJ see =PRS 'I know Jacqueline from a long time ago (lit. I am seeing Jacqueline for a long time.' (5) ipa dezoito ano 1SG.POSS eighteen.years ku COP =a =PFV =hè =LOC pèk= 3PL= walo arrive =ke =wei =n =INCEP =NMLZ =ERG eskola school =hè =LOC klaiwa e nonindigenous 3SG.POSS liplosi book tũlu ya= written 1SG= tapo know =maha =CONCS =lã =NEG =o =noi =MID.VOI =?? ũhu DEM =pèku =3PL =n =ERG ya= 1SG= miyawe be.smart apè= UNSP= =le =RES =ha =REC.PST2 =hei =3PL.SUBJ hila give =thehe Alto Mukaxai =when Upper Mucajaí 80 'When I was 18, those who arrived in Alto Mucajai, even though I didn't know the written word of the nonindigenous, those taught me how to be smart in school. (lit. those gave me wisdom)' (6) inaha thus ku =yalo Teus e =because God 3SG.POSS COP thã word =elihè ya= kẽa =LOC 1SG= work =wei sĩuthai ta= ku =o =lei =NMLZ a.little.bit VIS= COP =MID.VOI =REM.PST 'So I worked a little bit on the Word of God,' (7) ano year de of dois mil =elihè ipa dezoito ano two.thousand =DIR 1SG.POSS eighteen years ku COP =a =thehe ya= kẽa xuluku =layo =ma =PFV =when 1SG= work prepared =RES =REC.PST1 'In the year 2000, when I was 18 years old, I became prepare to work' (8) mahakii ya= miyawe =yo =ma CONCS 1SG= be.smart =RES =REC.PST1 'Yet, I became smart ' (9) Xakalina Isapeu Jacqueline Isabel Karoo Carole =pèku =3PL ya= miyawe 1SG.OBJ= be.smart =n =ERG professora Hosa female.teacher Rosa yalami many =ma =le =CAUS =RES =ma =he =REC.PST1 =3PL.SUBJ 'Jacqueline, Isabel, Carolina, teacher Rosa, many people made me smart.' (10) keteen liplosi tũlu ya= tapo =pè =haiki =le hoithehe so book written 1SG= know =UNSP =certainly =RES now 'Today I certainly know the written word.' (11) inaha ku =yalo hiyehè gravação =elihè ya= pè= ku =le thus COP =because here recording =LOC 1SG= UNSP= COP =PRS 'Thus I'm here at the recording.' (12) inaha ku thus COP ku =yalo cinquenta porcento =elihè yehek= apè= =because fifty percent =LOC 1DL.SUBJ= UNSP= =le =PRS 'That is why we (dual) are at 50%.' COP 81 (13) inaha thus =hè =LOC ku COP =yalo hiyehè =because here ya= 1SG= kẽa work =yalo yehek= apè= =because 1DL= UNSP= totihi good Teus God e thã 3SG.POSS word =yalo gravação ya= tha =because recording 1SG= do xiho be.happy =le =PRS 'Thus, because I am working here in God's good word, because I am doing the recording, we (dual) are happy. (Thus, we are happy because I am working here in God's good word, because I am doing the recording.)' (14) inaha thã apè= ku =o =he thus word UNSP= COP =MID.VOI =REC.PST2 'Thus was the word' 82 6.3 Appendix 3: Cassava drink (procedural) (1) Haiki =a certainly =PFV ‘ok (lit. certainly) (2) Ninam person =yeheku =1DL =n =ERG wãlo man =n =ERG haklak edible.yucca =uk =CLN:liquid tha =wei apè= nuhmuti =le do =HAB UNSP= unable =PRS ‘It is impossible for us (dual) ninam men to make haklauk.' (3) thuwe woman =n =ERG hutu farm =hè =LOC haklak edible.yucca =kok =CLN:root ha= SEQ= huku pull.out =li =RES =n =after haklak edible.yucca =uk =CLN:liquid apè= tha =wei UNSP= do =HAB 'After the woman pulls the cassava in the field, (she) makes the cassava drink.' (4) Hutu =hè kok= te =wei, farm =LOC CLN:root= grab =HAB ‘She grabs the root in the farm.’ (5) Yãno =hè kok= kãi= house =LOC CLN:root= SIM= ‘She returns home with the root,’ (6) kok= hõxi =la CLN:root= peel =PLU ‘She peels the root,’ =le =RES (7) kok= yalu =le CLN:root= wash =RES ‘she washes the root,’ =i =?? (8) kok= tu =ke =n waiha CLN:root= cook =INCEP =after after.waiting ‘after cooking, after waiting,’ (9) kama =n uk= apè= tha =wei 3SG =ERG CLN:liquid= UNSP= do =HAB ‘She makes the drink.’ kope =i return =?? =i =?? 83 (10) hoithehe ũhu now DEM tha do =le =RES =wei =HAB yalupu =kup two =CLN:two thèm day ha= hayi SEQ= pass =ki =INCEP =n =after ũhu DEM =thehe =when ninam person =pèk kẽa =3PL work =wei =NMLZ ku =a =PFV =thehe =when COP thuwe =n apè= nosiama =wei woman =ERG UNSP= be.in.charge =HAB 'Now, having done so after two days, when the people (ninam) work, the woman takes over.' (11) (12) mi ki NEG ãxko =wei drink =NMLZ 'It's not just drinking.' COP kẽa =wei work =NMLZ 'it is work.' ki =apè =UNSP =lã =NEG =i =?? COP (13) ai hutu =sik tha =kõ other farm =CLN:planted do =again ‘we (are going to) make another farm again.' (14) sik= CLN:planted= tha do =i =?? =kõ =again =o =xo =MID.VOI =COMIT =o =MID.VOI =lã =NEG =thehe =when ai hutu pemi =o =ithe other farm nonexistence =MID.VOI =FUT 'when (one) does not plant again, (one) will not have another farm (lit. there will not be another farm).' (15) (16) (17) inaha keteen kẽa thus SO work 'Because that's work.' =wei =NMLZ hei haklak =uk this edible.yucca =CLN:liquid 'This is the cassava drink.' inaha thuwe =n thus woman =ERG 'Thus the woman makes it.' apè= UNSP= apè= UNSP= ki COP apè= ku UNSP= COP ki COP =i =?? =wei =HAB =i =?? 84 (18) ku =thehe wãlo =when man 'Being that, all men go' COP (19) ũhu DEM =pèk =3PL kõmoli all =thehe thuwe pĩhalop =when woman husband apè= hu move UNSP= =e =n =3SG.POSS =ERG ũhu =wei =HAB DEM =thehe pèk =when 3PL kõmoli all ha= SEQ= naka call =li =RES =n =after ai other kõmoli all ha= SEQ= naka call =li =RES =n =after =pèk =3PL hutu =hèm pèk= kãi= apè= hu =wei farm =DIR 3PL= SIM UNSP= move =HAB 'Then, after the woman's husband has invited everyone, he invites others and they go simultaneously to the farm' (20) hemei ipa hutu let's.go 1SG.POSS farm 'Let's go, let’s make my farm.' (21) ipa thuwe 1SG.POSS woman tha do =a =PFV =xo =COMIT =n haklak =uk apè= =ERG edible.yucca =CLN:liquid UNSP= tha do =li =he =RES =REC.PST2 'My wife made the cassava drink' (22) inaha thus pèk= 3PL= ku COP =thehe =when kõmoli all apè= UNSP= hu move =wei =HAB 'Thus everyone goes.' (23) hutu farm =hè =DIR =ha =SEQ =kõ =ki =n =again =INCEP =after =mo =yo =REFLEX =RES pèk= 3PL= =n =after kẽa work pèk= 3PL= =wei =NMLZ pèk= 3PL= walo arrive pèk= 3PL= ha= =SEQ yalu wash ãxko drink =o =i =MID.VOI =?? 85 'After the workers arrive home from the farm again, after washing themselves, they drink.' (24) ũhu DEM =thehe kẽa =wei =when work =NMLZ =uk apè= =CLN:liquid UNSP= hipi give =a =DIST thuwe woman =n =ERG haklak edible.yucca =wei =HAB kẽa work =wei =NMLZ =pèk =hèm =3PL =DIR 'So the working woman give the drink to the workers.'