Constructional Polysemy and Mental Spaces in Potawatomi Discourse by Laura Ann Buszard B.A. (University of Michigan, Ann Arbor) 1991 M.A. (University of California, Berkeley) 1994 A dissertation submitted in partial satisfaction of the requirements for the degree of Doctor of Philosophy in Linguistics in the GRADUATE DIVISION of the UNIVERSITY OF CALIFORNIA, BERKELEY Committee in charge: Professor Richard A. Rhodes, Chair Professor Eve Sweetser Professor William Hanks Professor Leanne Hinton Fall, 2003 The dissertation of Laura Ann Buszard is approved: Chair Date Date Date Date University of California, Berkeley Fall, 2003 Constructional Polysemy and Mental Spaces in Potawatomi Discourse ©2003 by Laura Ann Buszard For James Abstract Constructional Polysemy and Mental Spaces in Potawatomi Discourse by Laura Ann Buszard Doctor of Philosophy in Linguistics University of California, Berkeley Professor Richard A. Rhodes, Chair This dissertation examines several grammatical features of Potawatomi, a Central Algonquian language, whose syntactic distributions in traditional narrative are different from those found in everyday discourse. These grammatical features include the verbal paradigmatic orders known as the independent and conjunct, a verbal prefix e-, and obviation. In everyday discourse, independents are main clause forms, and conjuncts are generally subordinate clause forms. The verbal prefix e- is a marker of factivity within a subordinate clause. In narrative, however, most main clause verbs take the conjunct prefixed with e-. The function of obviation in everyday discourse is largely syntactic, with several several obligatory contexts of application. In narrative, however, it is optionally used to foreground and background characters, and to represent shifts in viewpoint. These distributions raise the issue of the relationship between between syntactic structure and discourse structure, and present the challenge to linguistic theory of accounting for syntax that is dependent on discourse context. I argue that the discourse- dependent distributions of these grammatical phenomena can be explained in a cognitive 1 linguistic framework, which assumes that syntax is not autonomous, but part of a continuum of form / meaning pairings which includes the lexicon and discourse structures. Within this framework, I propose that the aspects of Potawatomi grammar described above participate in several constructions that map a particular grammatical form onto multiple functions in both syntax and discourse. Using Mental Spaces theory, I show that these functions are related to each other in the way they structure and index aspects of mental spaces networks. I also argue for a productive mental space blend in Potawatomi that takes as its input spaces syntactic and discourse uses of constructions. In this way, possible contexts for the application of a construction in one domain can be associated with established contexts in the other. When the cross-space mappings are made, the blend can be 'run' and the construction applied to the new domain. This blend demonstrates that a full semantic description of these constructions requires explaining their functions within the domains of syntax and discourse, as well the relationships between their functions across these domains. ACKNOWLEDGEMENTS This project would not have been possible without the help of a great number of people. My first debt of gratitude is to the elders who shared with me their knowledge of the Potawatomi language, and my friends in the various Potawatomi communities who helped support this research. In particular, I would like to thank Hap McCue, for sparking my interest in Neshnabemwen, and introducing me to Richard Rhodes. On Walpole Island, Reta Sands and her family, thank you for your guidance and kind hospitality during my first summer of research; and Dean Jacobs along with the staff of the Nin.da.waab.jig Heritage Center (I think I finally sold all of those T-shirts!). At the Pokagon Band, the dear people who taught me my first Potawatomi and who have since passed on, Julia and Martin Wesaw. In Hannahville, the elders in language class; Noreena Matrious and her family, for opening their home to me; the staff of the Visions Center, the staff of the Nah Tah Wahsh Language and Culture program, and Tom Miller. Carol Bergquist, for always being supportive of my research, and for encouraging me both in words, and by her example. Vicki and Donny Dowd, for their unwavering support and hospitality, and to Hannahville' s tribal council, for their dedication and support of Potawatomi Language revitalization. In Forest County, the Daniels family, particularly Sharlene White, who put me up and put up with me those long winter months. The elders of the Prairie Band in Kansas, particularly Jane Puckkee, Sarah Patterson, and Walter Cooper; and my friend and traveling companion, Suzanne Battese, and her delightful family. Above all, I must thank two elders in particular: Mary Daniels, for all of those months we worked together, for putting up with me asking for all 111 of those questions (I am still trying to figure out all of your answers!), and for telling wonderful, wonderful stories; and Jim Thunder, for his deep knowledge and love of this language, and for passing that on to his people, and to me. . . along with the accordion. For the elders who have passed on, I dedicate this work to your memory: Jack Sands, Julia Wesaw, Martin Wesaw, Bud On-Ja-Wa and Joe Migwanabe — I miss you all very much. At Berkeley, I would like to thank our Departmental staff; Belen Flores, Paula Floro and Esther Weiss, for keeping me from getting lost (or rescuing me when I did) in the wild world that is UC Berkeley. Bill Weigel, thank you for our conversations about this and other research, and for all you have done for me and for so many other Berkeley Linguistics graduate students — I don't think we could function without you! I would also like to thank my teachers, particularly my committee: Eve Sweetser, for her faith in my research (and for seeing the finished product in my messy drafts) and for introducing me to the theory of Mental Spaces; Leanne Hinton, for demonstrating the value of language revitalization, and for inspiring me in so many ways; and Bill Hanks, for always challenging me with a fresh perspective. Most of all, I must thank Rich Rhodes for introducing me to Potawatomi, and guiding me all these years; I always left your office in awe of your deep knowledge of language, and of Algonquian languages in particular. Lastly, and most importantly, I would like to thank my family: my mother and father for their love, and for inspiring in their children a love of learning; my brother Brad and his wife Michelle (I look forward to our future adventures!); my husband's parents for their love and selfless support; and most of all for my dear husband, James — with your love, all things seem possible. IV Constructional Polysemy and Mental Spaces in Potawatomi Discourse Table of Contents 1 . Introduction 1 2. Grammatical preliminaries 1 1 3. Theoretical preliminaries 32 4. Independents and conjuncts in everyday discourse 53 5. Verbal paradigms and mental space construction 72 in everyday discourse 6. Independents and conjuncts in narrative discourse 85 7. Mental space construction in narrative 118 8. Obviation in syntax and discourse 146 9. The obviation construction 187 10. Summary: Cross-domain mappings 232 Bibliography 246 Appendices A. Grammatical codes used in morpheme glosses 253 B. Glossed examples from Chapters 6 and 7 254 C. Narrative text: "Crane Boy" 284 1 Introduction 1.1 Background and goals The present study grew out of a descriptive problem in Potawatomi. The problem concerns the behavior of grammatical elements in discourse, and raises issues about the relationship between syntactic structure and discourse structure, and presents the challenge to linguistic theory of accounting for grammatical constructions whose distribution is dependent on discourse context. The problem, as I first encountered it, did not concern an unusual aspect of Potawatomi grammar, but one that is generally considered to be mundane: the distribution of main and subordinate clause verb forms. If one considers only sentences as found in everyday discourse, the problem is not apparent. It arises only in the context of comparing such sentences to those found in tradional narrative, where the distribution is quite different. In narrative, the majority of main clause verbs are marked as syntactically subordinate, and main clause verb forms 'proper' are used for special purposes: the speech of characters, background information, and the representation of narrative-internal viewpoint. The problem does not end here, however. There are other aspects of Potawatomi grammar whose 'syntactic' behavior does not match what one finds in narrative. One that is obvious from even a cursory glance at narrative is the use of a verbal prefix e- which is found on nearly every main clause verb form. In everyday discourse, however, its syntactic use is as a marker of factivity in a subordinate clause. 1 Another aspect of grammar with both syntactic and discourse domains of application is obviation. Obviation is a common feature in Algonquian languages that signals disjoint reference in third persons. Within phrases and clauses, obviation is obligatory: if there are two or more third persons, only one may be proximate; others will be obviative. In narrative, however, obviation can be used for stylistic purposes to foreground and background characters, and to represent point-of-view. These aspects of Potawatomi grammar raise the theoretical problem of accounting for grammatical constructions whose distribution is dependent on discourse context. A generative syntactic analysis, modular in its approach, would likely take individual sentences from everyday discourse as data, and might then state, for example, that independents are main clause verbs, and conjuncts are subordinate clause verbs. However, this analysis would founder if it were extended to narrative discourse, where the distributions of these verbal paradigms are quite different.^ The same is true for the preverb e-, which a syntactic analysis might describe as applying only to subordinate clause verbs. How then might one explain its proliferation to nearly every narrative main clause verb? However, the modular approach to syntax has probably been most detrimental to the study of obviation in Algonquian languages, where studies commonly ^ Interestingly enough, the primary description we have of Potawatomi (Charles Hockett's work from the Structuralist era), has the opposite problem of taking narrative discourse as its basis for syntactic description (Hockett's data primarily came from traditional narrative texts), thereby missing most of the interesting behavior of these grammatical phenomena in everyday conversational discourse. 2 discount its discourse use as outside the scope of syntactic study, or even the domain of linguistic inquiry.^ While the primary goal of a modular approach to syntax is to capture generalizations about well-formedness, seen in a less forgiving light, it can only account for the well-formedness of a part of the grammatical constructs speakers are capable of generating. My assumption in the present study is that we have more to gain from studying the function of grammatical phenomena in both syntax and discourse than from excluding the data from either domain, a priori, from our analysis. I will argue that the behavior of discourse-sensitive morphosyntax in Potawatomi is principled, and moreover makes use of mechanisms already needed to explain sentence-level structures. My goal is to show that the syntactic behavior of these grammatical phenomena in everyday discourse and their textual use in traditional narrative discourse are related to each other, and will propose a model that captures these relationships. 1.2 A cognitive approach The theoretical approach taken here is Cognitive, that is, it rests on the assumption that language is not a separate, isolable, faculty of the human mind but is intimately bound up with general cognitive processes involving perception, processing, reasoning and construal. Two theories developed within this overarching framework are central to this study: Construction Grammar and Mental Spaces. These theories, along with their notational conventions, are presented in Chapter 3 . Construction Grammar, which is a unificational theory of syntax, will generally be used for syntactic analyses. I ■ For an example of this approach, see Aissen (1997). 3 will allude to the theory in Chapters 4 and 6 when I present the Conversational Construction and Narrative Construction, respectively, but the theory is heavily utilized in Chapter 9 for the representation of obviation. The second. Mental Spaces theory, is not strictly speaking a theory of discourse (it was developed, in part, to address problems of reference) however it has proven to be very useful in the analysis of narrative. Its advantage in the present study is it provides a means of distinguishing narrative and everyday discourse. The theory of Mental Spaces figures prominently in Chapters 3, 5, 7, and 9, and 10. 1.3 Data It is often difficult to obtain data from different discourse genres from published sources; a particular problem for those working on endangered languages who often rely on philological work to help fill in gaps where there has been grammatical attrition. For example, it is clear from Hockett's sketch of Potawatomi that narrative discourse formed the basis of his grammatical description. With regard to the uses of paradigmatic orders outside of narrative, we learn only that the independent is used "for statements and some questions in ordinary conversation" and that the conjunct is used "in certain types of dependent clauses" (Hockett, 1948a, p. 9). Given that the distributions of the paradigms are very different in these two discourse types, it is surprising that non-narrative discourse received no further attention. This omission was likely due to the fact that his data consisted primarily of narrative texts. It has been the tradition among Americanists, especially where field time is limited, to primarily elicit narratives, and within this type of discourse, the even narrower genre of mythological text. What is more unfortunate about the lack of conversational data, is the American Structuralists were working at a time when many speech communities were still robust, and conversational data would have been easier to obtain (although, not necessarily easier to record in a notebook) . Charles Hockett conducted his research on Potawatomi about the time most speakers shifted to English, and they would raise their children as first language English speakers. Today, there are a very few elderly fluent speakers left. Most do not use Potawatomi extensively in the home, because their children and grandchildren do not speak or understand it. In many cases, speakers do not live very close together. As a result of these factors, conversational data is rather difficult to obtain, and, admittedly, I collected very little of it myself In the process of using Hockett' s materials as a basis for elicitation and comparison, however, I noticed that the morphosyntax of the modem elicited data was quite different from that recorded in traditional narrative. Upon examination, the primary difference turned out to be with respect to narrative clauses — the reported speech of characters in narrative matched the elicited data. When I began working with speakers to create pedagogical materials, I found that the morphosyntax of their constructed conversations matched those of the elicited data. For these reasons, I would not go as far as to say that the elicited data I have used is conversational /?er se, but I believe that with respect to the linguistic parameters I am examining, it is good representation of the type of language used in everyday discourse. The data used for this study comes from several sources. The narrative data was largely collected by Hockett in the 1940's, which I have been in the process of translating. It is currently unpublished except for the two texts included in his UAL series on Potawatomi (Hockett, 1948b). The examples cited here come from about ten of these narratives, the majority of which were told by Jim Spear, and a few by his wife Alice Spear (one of her narratives "Crane Boy" is provided in Appendix C). These are cited either as JS (Jim Spear) or AS (Alice Spear). For this subset, I am reasonably satisfied with the glosses and free translations. Other examples come from narratives told to me during the period of 1994 - 1996 by a conservatively fluent female speaker, cited as MD. For data on everyday discourse, I include the elicited examples in my own fieldnotes, which are cited as POEX. Those examples annotated JTNB are taken from the conversations in a pedagogical workbook developed by fluent speaker Jim Thunder with Kim Wensaut (1998). 1.4 Chapter organization The structure of the text is as follows. Chapter 2, 'Descriptive preliminaries' provides a background for the grammatical topics to be addressed in later chapters. Chapter 3, 'Theoretical preliminaries' presents the Cognitive orientation of the analysis, and Construction Grammar and Mental Spaces theory. In this Chapter, I also argue for an elaborated representation of ground in the Mental Spaces theory. This representation will become important for contrasting various types of information in traditional narrative. Chapters 4-9 are arranged in pairs, with a descriptive chapter followed by a theoretical chapter. My intent in using this type of presentation is twofold; first, to make the descriptive information as accessible and theory-free as possible. Since the descriptive topics here have not been significantly addressed for Potawatomi, and in some cases Algonquian languages in general, I feel it important to give them due 6 attention. Secondly, I did not wish to encumber the line of theoretical argumentation with excessive descriptive detail. Each of these chapters is summarized in more detail below. Chapter 4 addresses the use of two types of verbal inflections in Potawatomi, the independent and conjunct, along with a preverb e-, as they are used in everyday discourse. It is shown that independent verbs are used for main clauses, and conjunct verbs are generally used for subordinate clauses. The preverb e- is shown to to be a marker of factivity in subordinate clauses. However, there are a few contexts where a conjunct can occur in a main clause, particularly when accompanying one of several particles indicating speaker evaluation. In addition, conjuncts can occur in main clauses without an accompanying particle if this evaluation is available in the context. I argue that these evaluations provide a context of subordination, which is satisfied by the use of the conjunct. This pattern of main clause independents, subordinate clause conjuncts and the preverb e- is introduced as the Conversational Construction (CC), which will be contrasted with the pattern of these grammatical elements in narrative.'' In Chapter 5, 1 present a Mental Spaces theory analysis of the elements of the Conversational Construction. I argue that in their everyday uses, independents structure Space R (the space which represents the "Reality" domain), and conjuncts always structure a space that is embedded in Space R. The preverb e- is a marker of factivity of ^ The linguistic entities I am referring to here as constructions are complex, in that they have analyzeable pieces which are themselves constructions. When these subconstructions combine, they contribute elements of their semantics to the larger 'super' construction. 7 an embedded space. These basic uses are contrasted with the function of these grammatical elements in narrative in Chapter 7. Chapter 6 presents an analysis of independents, conjuncts and the preverb e- in traditional narrative discourse. I argue that the use of main clause conjuncts is the basic narrative pattern which reflects narrative foreground. I call this basic narrative pattern the Narrative Construction (NC). By contrast, the use of main clause independents (that is, the Conversational Construction) in narrative reflects background information, either settings, explanations, or evaluations. Other uses of the Conversational Construction reflect a narrative-internal perspective, or viewpoint, which is used for direct speech, vividness, epistemic distance, or semantic opposition. I argue that the several uses of narrative-internal viewpoint probably arose out of the use of the Conversational Construction for direct speech. Chapter 7 presents a Mental Spaces theory analysis of the use of the Narrative and Conversational Constructions in traditional narrative discourse. I argue that the use of the Narrative Construction reflects that narrative is generally set up as an embedded network within a larger non-narrative discourse. The use of the Narrative Construction to mark foreground is metonymic for narrative discourse as a whole. When the Conversational Construction is used in narrative, it always indexes its basic use in the "Reality" domain in some way. With respect to background information, there is a contextual focus on one of the discourse participants. The various uses of the Conversational Construction for internal viewpoint reflect that viewpoint is inside of the focused narrative domain, whereas an external viewpoint (represented by the use of the Narrative Construction) reflect that viewpoint is outside of the focused narrative domain. 8 Chapter 8, contains a description of the use of obviation in Potawatomi. I describe both the marking of obviation on nouns and verbs, as well as the syntactic contexts for obviation. I then argue, by analyzing a traditional narrative, that the appearance of Potawatomi as a largely syntactic obviation language is due to a separate treatment of transitive and intransitive main clause verbs. Intransitive verbs reflect the syntactic pattern of obviation, whereas transitive verbs reflect the use of a hierarchical ranking of discourse participants. I show that despite this tendency towards syntactic obviation, the narrator is clearly working to maintain the main character as proximate, and makes use of discourse obviation in some very subtle and interesting ways. I argue that a possible path for a discourse obviation language to become a syntactic obviation language is grammaticalizing proximates as subjects of main clause intransitive verbs of speech, and that Potawatomi shows this change in progress. In Chapter 9, 1 analyze the use of obviation in terms of Construction Grammar and Mental Spaces theory. I argue that the various uses of obviation in syntax and discourse reflect the use of a basic obviation construction that ranks multiple third persons, and assigns proximate status to the highest ranked third person. Various "instance" constructions that inherit the obviation construction provide the details of specific ranking schemes."^ I also show that the use of discourse obviation, in particular proximate shifts, can be accomodated by associating particular nominal rankings to various viewpoints in the Mental Spaces theory networks. These networks are then indexed inside of particular obviation instance constructions. '' The term instance construction is from Goldberg (1995), and will be explained m more detail in Chapter Chapter 10 concludes with a discussion of Mental Space blends in Potawatomi discourse. I argue that independents, conjuncts, the preverb e-, and obviation reflect a productive blend in Potawatomi that takes as its input spaces syntactic and discourse uses of constructions. In this way, possible contexts for the application of a construction in one domain can be associated with established contexts in the other. When the cross- space mappings are made, the blend can be 'run' and the construction applied to the new domain. I argue that the existence of these blends demonstrates that a full description of these constructions requires predication in multiple grammatical domains, syntax and discourse. There are also three appendices: Appendix A contains a list of the grammatical codes used in morpheme glosses. Appendix B provides interlinear glosses of the textual examples used in Chapter 6 and 7. Appendix C contains the narrative "Crane Boy", which is discussed in Chapters 8 and 9, presented with interlinear glosses and facing translation. 10 Aissen, Judith. 1997. On the Syntax of Obviation. Language, 73.705-50. Goldberg, A. 1995. Constructions: A Construction Grammar Approach to Argument Structure: Cognitive Theory of Language and Culture. Chicago: University of Chicago Press. Hockett, Charles. 1948a. Potawatomi L Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi IV: Particles and Sample Texts. International Journal of American Linguistics, XIV. 213-25. Thunder, Jim Sr. and Kim Wensaut. 1998. Bodewadmimwen Nswe Mbok (Bodewadmi Language Book Three): Jim Thunder and Kim Wensaut, publishers. 11 2 Grammatical Preliminaries 2.1 Introduction The purpose of this chapter is to introduce those aspects of Potawatomi grammar which will be addressed in later chapters, and to provide a background for understanding the system of transcription and interlinear annotations. It is therefore not intended as a grammatical description, or sketch. For a fuller description of Potawatomi grammar, particularly phonology and morphology, the reader is referred to Hockett's series of articles on Potawatomi in the InternationalJournal of American Linguistics (1948a; 1948b; 1948c; 1948d).^ 2.2 Background on Potawatomi Potawatomi is the heritage language of the Potawatomi people, who are indigenous to the Great Lakes region of North America.^ In Potawatomi, the language is sometimes refered to as Bodewadmimwen ('the language of the Potawatomi s'), or more commonly as Neshnabemwen ('the language of the people'). It is an Algonquian language, of the Central branch, which includes other languages such as Ottawa, Ojibwe, Cree, Fox, Shawnee and Miami. Its closest linguistic relatives are Ojibwe and Ottawa, although this ^ This series of articles is a revision and distillation of the material in his dissertation on Potawatomi (1939). The article series cited above is more readable, and generally easier to obtain, than the dissertation. " Today, largely as a result of 19''' century U.S. government relocation policies, Potawatomi people live on or near reservations across the Midwestern United States, and in adjacent areas of Ontario, Canada. 11 is somewhat obscured by vocabulary and grammatical changes resulting from an extended period of contact with Fox speakers.' Potawatomi is a polysynthetic language. It is 'pro-drop' in that verbal participants are represented by verbal inflections, which may then further specified by NPs. Along with being pro-drop, it is also non-configurational in that word order is generally flexible, and governed by discourse principles."^ Potawatomi grammar is probably best known among linguists for its system of inflections in the independent paradigm, particularly on transitive verbs, which has been frequently used to demonstrate the robustness of various morphological theories.^ While Potawatomi is certainly interesting in this respect, it should be noted that many Algonquian languages have similar paradigms, and equally complex systems of verbal inflection. 2.3 Guide to the orthography -.6 The orthography used here is known as the WNALP system, and was developed in the 1970's by a team of native speakers and linguists. It is a phonemic system. ^ More precisely, Sauk speakers. There are differences between Fox and Sauk, however the differences are irrelevant for the present discussion. I will generally cite Fox because of the availability of lexical materials in that language. '' There are some word order restrictions however; such as the placement of second-position particles and the negative particle jo which precedes the verb. ^ For a fairly typical example, see Anderson (1992). "^ Wisconsin Native American Languages Program. 12 designed for the purposes teaching Potawatomi as a second language. The orthographic representation of phonemes is fairly straightforward, and is given in the chart in (1). 13 (1) ORTHOGRAPHY CHART Consonants Vowels Orthographic Phonemic Orthographic Phonemic b /b/ a /a/ P /p/ e / / d /d/ i /i/ t III /o/ g l%l e / / k /k/ 5 / / m /m/ n /n/ w /w/ y /y/ s /s/ z /z/ sh / / zh / / h /h/ ch /t / J /d / 14 There are, in addition, a few special symbols that are used in the morphophonemic representations in interlinear glosses. These are described below in Section 4. 2.4 Morphophonemic processes and representations A couple processes important for morphophonemic representations are noted here, as well as the set of morphophonemic symbols used in glosses. Final devoicing. Voiced consonants are devoiced in word-final position. The voicing resurfaces when suffixes are added. The following are a few examples, showing the alternation between stem-final consonants in singular and plural forms: (2) FINAL DEVOICING SINGULAR PLURAL mtek 'tree' mtegok 'trees' mskogat 'yam belt, sash' mskogadek 'yam belts, sashes' nnech 'my hand' nnejen 'my hands' Weak vowel deletion. All instances of /e/ and some lol vowels are subject to a process of deletion. These are known as weak vowels, and are represented by {E} and {O}, respectively in morphophonemic forms. '' In a sequence of weak vowels, odd vowels delete (counting from the beginning of the sequence). If the sequence is ^ This rule basically affects Proto-Algonquian (PA) short vowels. Potawatomi merged short PA *i and *a to schwa (/e/). There was also a merger of PA short *o and long *o:, with the result that some lol vowels delete, and some do not. The term 'weak' vowel is from Hockett (1948a). 15 interrupted, the count begins again with the first sequential weak vowel. Final weak vowels are not subject to deletion. This process is illustrated below in (3). Weak vowels are numbered, and long vowels are represented by "L". Vowels preserved in final syllables are shown with parentheses surrounding the number. Note that prefixes can contribute short vowels, resulting in different pattern of deleted stem vowels as compared to base forms. (3) WEAK VOWEL DELETION < b E kw ezhEgEn> -^ bkwezhgen 'bread' 1 L 1 (2) < n E - b E kw ezhEgEn-Em> -^ nbekwezhgenem 'my bread' 1 2 L 1 2 (3) Palatalization. The remaining morphophonemic symbols used are for those consonants that participate in a process of palatalization, where morpheme final /n/, /d/, III and /s/ become /zh/, /j/, /ch/ and /sh/, respectively, before a morpheme initial Id or I'll. These consonants are represented by the capitol letters {N}, {D}, {T} and {S}. 2.5 Parts of speech The parts of speech are noun, verb, pronoun, prenoun, preverb, and particle. Nouns and verbs are subject to inflection; these are described in more detail in Sections 6 and 7. The remaining parts of speech are introduced here. 2.5.1 Pronouns Potawatomi has two primary pronoun series, personal pronouns and demonstratives. The set of personal pronouns is shown in (4). Because nouns and verbs 16 make anaphoric reference in their inflections and do not require the use of pronouns, the function of the free pronouns is essentially for emphasis. (4) PERSONAL PRONOUNS POTAWATOMI ENGLISH GLOSS nin I gin you win he, she ninan we (excluding the addressee) ginan we (including the addressee) ginwa you (plural) winwa they The set of demonstrative pronouns has three series; proximal, medial and distal. In discourse, the demonstrative in the medial series have a determiner-like function. Each series is given below: (5) PROXIMAL DEMONSTRATIVE PRONOUNS ANIMATE INANIMATE SINGULAR ode 'this' PLURAL gode 'these' node 'this (obviative), these' 17 (6) MEDIAL DEMONSTRATIVE PRONOUNS ANIMATE INANIMATE SING. ow (frequently reduced to o) 'that' iw (frequently reduced to /') 'that' PLURAL giw (frequently reduced to gi) 'those' niw (frequently reduced to ni) 'that (obviative), those' (7) DISTAL DEMONSTRATIVE PRONOUNS ANIMATE INANIMATE SINGULAR ago 'that over there' e'i 'that over there' PLURAL egi 'those over there' eni 'this (obv.) over there, those over there' 2.5.2 Prenouns and preverbs Prenouns and preverbs are prefixes that attach to nouns and verbs, respectively. They attach directly before the stem, so if personal prefixes are used, they will attach to the preverb or prenoun (in the case of possession). Initial change (see Section 7) will affect the first preverb, if there are any. Each is a rather small set, consisting of less than 100 forms. Because prenouns and preverbs behave phonologically as separate words, they are written with a following hyphen. While most are fairly productive, their use is often specialized (for example, msko-bneshi, literally 'red-bird', refers to a Cardinal, and not just any red bird). The most important for our discussion are the preverbs, which include the tense preverbs (see Section 7), as well as the factive preverb e-. Ordering restrictions among 18 preverbs require that e- be first in the preverb sequence, followed by any tense / modal preverbs. 2.5.3 Particles The set of particles in Potawatomi is a closed class consisting of approximately 300-350 lexemes. Commonly used particles are 70 'no, not', ne 'question', and zhe na 'emphatic'. The category includes numbers, exclamations, words with discourse functions and a large set of words with adverbial and adjectival meaning, such Zi^jayek 'air, megwa 'still, yet', gnebech 'maybe'. Although particles do not inflect, they can be morphologically complex. Many take prenouns and preverbs, and have inflectional suffixes such as the locative azhodakik 'over the hill' or the dubitative jv^tfeA: 'must be, maybe, I wonder'. Particles can also occur as groups or clusters. Where this is the case, they are semantically idiomatic, and behave syntactically like a unit. 2.6 Categories of inflection Person. The categories for person are first, second and third. Within first person, there is a distinction between inclusive 'we' (that includes the addressee) and exclusive 'we' (that does not include the addressee). Human gender is not distinguished in third person pronouns or inflections. These can be illustrated with the series of personal pronouns, shown in (4) above. Obviative. Within third person, there is a distinction between proximate and obviative: In contexts where there is more than one third person, only one third person may be proximate, and any other third person will be obviative. There are both syntactic and discourse contexts for obviation. Some syntactic contexts are obligatory. 19 Obviatives are marked with inflections, proximates are unmarked. Nouns inflect for obviation, and verbs inflect for obviative agreement. Only animate nouns inflect for obviation, however inanimate nouns can trigger obviative agreement on verbs. The following shows the obviative inflection that is obligatory on animate possessees with third person possessors: (8) OBVIATION OF POSSESSEE ngwes 'my son' ggwes 'your son' wgwesen 'his / her son (obv.)' Number. The numbers are singular and plural. Gender (animacy). The grammatical genders are animate and inanimate. The animate category includes notional animates; items of cultural / religious importance such as sema 'tobacco' and dewe 'gen 'drum'; some objects that move without the apparent application of external force such as dabyan 'automobile', gizes 'sun', negos 'star'; and other non-notional animates such as kek 'kettle' or mjenkawnek 'mittens'. The inanimate category includes everything else. Other nominal forms . Most nouns additionally have locative, diminutive and pejorative forms. Some kinship terms have vocative forms. Nouns can also take the verbal preterite ending -ben to mean 'former' or 'deceased', as in nosben 'my late father'. Order. There are three orders: independent, conjunct, and imperative. In everyday discourse, independents are used in main clauses, and conjunct verbs in subordinate 20 clauses. Conjunct verbs are used in main clauses in content questions, and with certain 'subordinating' particles (these uses are described in more detail in Chapter 4). The primary inflectional difference between the two orders, are that the independent has both personal prefixes and suffixes, whereas on the conjunct, person markings are strictly suffixal. This is shown in (9), with the intransitive verb stem majit 'leave' :^ (9) INDEPENDENT AND CONJUNCT ORDERS COMPARED INDEPENDENT CON.TUNCT I nmaji majiyan you sg. gmaji majiyen he maji(wak) majit obviative majin majinet we exclusive nmajimen majiyak we inclusive gmajimen majiygo you pi. gmajim majiyek they majik majiwat Imperatives are used for commands as well as polite requests. The verbal orders are illustrated in (10) below using the verb bidget 'enter; come in': (10) VERBAL ORDERS COMPARED INDEPENDENT Bidge. 'He is coming in.' CON.TUNCT (gishpen) bidget '(if) he comes in' IMPERATIVE Bidgen! 'Please, come in.' Verbs are cited in the conjunct indicative form for third person subject. 21 The conjuct occurs in two forms, plain and changed. Changed conjuncts have ablaut of an initial vowel, which is known as initial change. The form of initial change is shown in (1 1) below: (11) INITIAL CHANGE PLAIN CHANGED {i} a {E} e {0} wa {0} we {a} (no change) {e} (no change) Examples are shown in (12) below, comparing the plain conjunct with participle. The morphophonemic form underneath the plain conjunct form shows the presence of weak vowels affected by ablaut, which may be deleted in the inflected form. 22 (12) EXAMPLES SHOWING INITIAL CHANGE IN PARTICIPLES CONJUNCT GLOSS PARTICIPLE GLOSS minket {minEked} 'if he picks berries' manket 'the one who picks berries' bmoset {bEmOsed} 'if he walks' bemset 'the one who walks' bod wet {bodEwed} 'if he builds afire' bwadwet 'the one who builds a fire' wdemat {OdEme/ad} 'if he/she smokes' wedmat 'the one who smokes' majit {majid} 'if he/she leaves' majit 'the one who leaves' deb sat {debEsad} 'if he has enough' deb sat 'the one who has enough' Changed conjuncts are generally found in contexts of presupposition (such as the main clauses of wh- questions and certain adverbial clauses), whereas plain conjuncts are found in hypothetical or irrealis contexts. Participles are identical in inflection to the changed conjunct, except for third person obviative and third person plural forms. Examples of conjunct forms are shown below in (13), using the verb bidget 'come in; enter' : 23 (13) VERBAL ORDERS AND INITIAL CHANGE CONJUNCT, PLAIN (gishpen) bidget '(if) he comes in' CONJUNCT, CHANGED ga-bidget 'after he came in' CONJUNCT, PARTICIPLE badget 'the one who comes in' Mode. The major verbal modes are the indicative, negative, preterite, dubitative, negative preterite, and negative dubitative. These are found as suffixal inflections. The independent mode has an indicative, negative, preterite, dubitative, negative preterite and negative dubitative mode. Independent negative modes require the use of a preposed negative particley'o. The conjunct has an indicative, preterite and dubitative, but does not have negative modes. Negation is expressed on conjunct verbs with the use of a preverb bwa-. The imperative order has a prohibative mode. The function of each mode is illustrated below with the independent form of the verb wabmat 'see someone' : • Indicative: the basic form of an assertion or yes/no question. Gwabma. 'You see him' . Gwabma ne? 'Do you see him?' • Negative: requires the negative parti cleyo in addition to negative suffixal inflection: Jo nwabmasi. 'I don't see him.' • Preterite: expresses something that happened habitually in the past, often accompanied by the particle neko 'used to' : Neko nwabma ben . 'I used to see him' 24 • Dubitative: in the present tense, the dubitative expresses doubt; in the past tense, it expresses an inference: Gnebech nwabma dek . 'Maybe I see him.' Ngi-nwabmadek. 'I must have seen him.' Combinatory modes: these require the use of the negative parti cleyo: o Negative preterite: J(i nwabma siben . 'I didn't used to see him.' o Negative dubitative: in the present tense, often used with the particle gnebech 'maybe' : Jo gnebech nwabmasidek. 'I might not see him.' In the past, expresses an inference: Jo ngi-wabma sidek . 'I must not have seen him.' The modes are illustrated below with the intransitive verb majit 'leave' : (14) VERBAL MODES INDEPENDENT INDICATIVE Maji. 'He is leaving.' INDEPENDENT NEGATIVE Jo maji si. 'He is not leaving.' INDEPENDENT PRETERITE Majiben. 'He used to leave.' INDEPENDENT DUBITATIVE Majidek. 'He supposedly left.' INDEPENDENT NEGATIVE PRETERITE Jo majisiben. 'He did not leave.' INDEPENDENT NEGATIVE DUBITATIVE Jo majisidek. 'He must not have left.' Tense. The tenses are past, present (unmarked) and future. Present tense is unmarked; past and future tenses are marked by verbal prefixes. There are two future tenses, one of which is volitional: 25 (15) TENSE INFLECTION PRESENT Maji. 'He is leaving.' PAST Gi-maji. 'He left.' FUTURE Ge-maji. 'He will leave.' (he is going to) FUTURE VOLITIONAL Wi-maji. 'He will leave.' (he wants to, or intends to) Transitivity. Potawatomi verbs inflect for the presence or absence of an object. If the verb has an object then the verb will further inflect based upon the animacy of the object. If the verb does not take an object it will inflect for the animacy of the subject. These parameters result in the following division of verbs into four main categories of inflection, or paradigms: transitive animate (TA), transitive inanimate (TI), animate intransitive (AI), and inanimate intransitive (II). 2.7 Nominal Inflection Noun stems are either animate or inanimate, as described in (Section 5) above. Nouns inflect for number and obviation. Possessed nouns inflect for the person and number of the possessee. Plural inflection. Animate and inanimate nouns have different plural inflections. Animates take the plural ending /-k/, and inanimates take the plural ending /-n/, as shown by the following (a 'connective' /e/ is added before the plural ending if morphophonemic form of the noun stem ends in a consonant, I usually show this connective as part of the suffix in morphophonemic representations): 26 (16) PLURAL INFLECTION OF ANIMATES AND INANIMATES ANIMATE PLURAL INANIMATE PLURAL nene 'man' {EnEnE} nenwek dopwen 'table' {dopEwEn} dopwenen mtek 'tree' {mEtEg#0} mtegok mkezen 'shoe' {mEkEzEn} mkeznen mdamen 'corn, kernel of com' {mEdamEn} mdamnek bkwezhgen 'bread' { bEkwezhEgEnEn } bkwezhgenen Obviative inflection. Obviation is marked only on animate nouns, however verbs show agreement with obviative inanimates. Obviative nouns are inflected like inanimate plurals, taking the suffix l-nl as shown in the following table: (17) INANIMATE PLURAL AND OBVIATIVE INFLECTION COMPARED SINGULAR PLURAL OBVIATIVE ANIMATE wabozo 'rabbit' wabozoyek wabozoyen dabyan 'car' dabyanek dabyanen INANIMATE dopwen 'table' dopwenen N.A. mkezen 'shoe' mkeznen N.A. Possessive inflection. Many nouns show the use of the possessive suffix {Em}. In addition, possessed nouns inflect for the person and number of the possessee: 27 (18) POSSESSIVE INFLECTION MORPHOPHONEMIC FORM POTAWATOMI WORD ENGLISH GLOSS { nE-bnEakwan-Em } nbenakwanem 'my comb' { gE-bEnakwan-Em } gbenakwanem 'your comb' { wE-bEnakwan-Em } wbenakwanem 'his/her/its comb' { nE-bEnakwan-Em-Enan } nbenakwanmenan 'our (excl.) comb' { gE-bEnakwan-Em -Enan } abenakwanmenan 'my (incl.) comb' { gE-bEnakwan-Em -Ewa } gbenakwanmewa 'your (pi.) comb' {wE-bEnakwan-Em-Ewa} wbenakwanmewa 'their comb' 2.8 Verbal inflection Basic information about the inflection of AI, II, TI and TA verbs is provided below. For detailed example paradigms, the reader is referred to Hockett (1948c). Animate Intransitive. AI verbs inflect for the person, number, and obviation of the animate subject. Inanimate Intransitive. II verbs inflect for person, number, and obviation of the inanimate subject. II's optionally take a suffix {-mEgEd}, which is known as an augment. This morpheme directly follows the stem and is then followed by inflections. Transitive Inanimate. TI verbs inflect for person, number and obviation of the subject, and optionally for number of the primary object Transitive Animate. TA verbs inflect for person, number and obviation of the subject and primary object. TAs have an inverse system involving first, second and third persons. The inverse system indicates whether the personal prefixes are the properties of 28 the subject or primary object. The person hierarchy used for this system is second person > first person > third person. If the subject is higher on this hierarchy than the primary object, a direct theme sign will be used. If the subject is lower than the primary object, an inverse theme sign will be used. There are four theme signs, two each for direct and inverse, depending on whether a third person is involved (non-local) or is not involved (local). The theme sign directly follows the stem and is indicated in interlinear glosses: ^ (19) TA THEME SIGNS DIRECT INVERSE Local (not marked) {En} Non-local {a} {EgO} 2.9 Interlinear glossing conventions Examples in the following chapters are cited in a few different ways. If it is necessary to discuss a particular morpheme, examples are given a three-line interlinear Hockett describes the direct, local theme sign as morphophonemically zero, but that it causes palatalization. Sources for Ottawa such as Rhodes (1976) and Valentine (2001) cite this theme sign as /-i/, the Potawatomi equivalent of which would be {E}. Hockett's decision may have been based on abstractness, since the {E} is never found in forms — only its effects may be observed. Since morphophonemic forms are already abstract, I agree in principle with Rhodes and Valentine, however follow Hockett's practice here. The inverse, local marker has two forms: {En} which is used in the conjunct paradigm, and {EnE} which is used in the independent. There are inconsistencies m Hockett's treatment of the inverse, non-local theme sign. I follow the practice of using {EgO} with final short {O}, rather than the more standard {Egw}. 29 gloss with a free translation, as shown below in (21). Line 1 contains the Potawatomi example, transliterated into the WNALP phonemic writing system. Line 2 divides words into their component morphemes, represented in morphophonemic form. This line shows a division into stems and inflectional morphemes; but generally does not show stem- internal derivational morphology. Line 3 contains morpheme glosses (a key to grammatical gloss abbreviations is given in Appendix A), and Line 4 contains the free translation. (20) FOUR LINE GLOSS Line 1 Line 2 Line 3 I me se ngodek neshnabek e-wdodanwat iw mE sE nEgOd-Eg EnEshEnabe-g e- wEdodanE -wad that.INAN EMPH EMPH one -LOG person -PL FCT-have . a . village\AI-35 . C Line 4 : Once there was a village, and someone was destroying their gardens and wells. When the focus is not on an individual morpheme or morphemes, three line glosses are given, omitting the morphophonemic line. (21) THREE LINE GLOSS E-nme-gisekwet mdemoze, FCT-getting.to.be-finish. cooking\AI=3 . C old .woman e-bye-bidgeshkak gche-emkwan. FCT- come -enter .with . body\II=0 . C big- spoon When the old lady is almost through cooking, in comes a big spoon. (AS: 1:3:51) When longer stretches of Potawatomi are given, my convention is to present an annotated facing translation, as in (23). Here particular constructions in the Potawatomi sentences are annotated with brackets and subscripts (these are described in Chapter 6). Underlined verbs on the Potawatomi side roughly correspond to the underlined verbs on the English side (where the order of the verbs in the translation does not match the 30 Potawatomi, subscript indices are given). Lines from the text are given along the left hand side, and the code for the text cited is given below the English translation. (22) FACING TRANSLATION 44 [Iw je e-bme-byat niw beshkmwen e-nat]uc [ "Nseze ! Gyetnam nzeges . ] cc 45 [ Nwebi ' we . ] cc 46 [Weye zhode nshiwnagze anwe ge gin gne s hi wnagwe s nesh je win nwech.]cc 47 [Ibe gge- zhyamen ; geten nshiwnagze ■ " ] cc 48 [Beshkmwe e-kedot , ] hc [ " Gzhyamen , gge - we - wabmamen . " ] cc When he [Rabbit] came across the hon he said to him, "Brother, I'm very scared . I'm running away from someone. Someone here is pretty scary ; and you're scary , but he's even worse. Let's go over there; he sure is scary ." Lion said, " Lets go and take a look at him." (JS.4.1) If the two line gloss or facing translation is used, the full interlinear gloss is available in either Appendix B (sentential examples) or Appendix C ('Crane Boy' narrative). 31 Anderson, Stephen R. 1992. A-Morphous Morphology .vol. 62: Cambridge Studies in Linguistics. Cambridge: Cambridge University Press. Hockett, Charles. 1939. The Potawatomi Language, Anthropology, Yale University: Ph.D. dissertation. — . 1948a. Potawatomi L Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi II: Derivation, Personal Prefixes, and Nouns. International Journal of American Linguistics, XIV. 63-73. — . 1948c. Potawatomi III: The Verb Complex. International Journal of American Linguistics, XIV. 139-49. — . 1948d. Potawatomi IV: Particles and Sample Texts. International Journal of American Linguistics, XIV. 213-25. Rhodes, Richard A. 1976. The Morphosyntax of the Central Ojibwa Verb, Linguistics, University of Michigan: Ph.D. Valentine, J. Randolph. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press. 32 3 Theoretical Preliminaries 3.1 Introduction: A cognitive approach The analysis which will be presented in this study is fundamentally cognitive. That is, I assume that language is not a separate, isolable, faculty of the human mind but is intimately bound up with general cognitive processes involving perception, processing, reasoning and construal, and that our theories must take these processes into account. With respect to grammar, I assume the following^ • Grammar is inherently symbolic, involving form-meaning pairings of phonological material with semantic structure. Syntax, in particular, is not modular or autonomous, but is part of a continuum that includes the lexicon, morphology, syntax, and (I would argue) discourse. As form-meaning pairings, syntactic constructions, like lexemes, exhibit semantic polysemy, where a single grammatical form is associated with multiple related senses. And, like lexical polysemy, multiple related senses are expected as the norm for constructions. ^ These principles reflect several sources in the cognitive literature, including Langacker (1991) , Goldberg (1995), and Lakoff (1987). 32 The cognitive linguistic theory which is most central to the theoretical discussion in this study is Mental Spaces. Mental Spaces theory will generally be used for the representation of discourse, and figures prominently in Chapters 5, 7, 9, and 10.^ Another cognitive theory, Construction Grammar, will generally used for syntactic representations. This theory is alluded to in the presentation of two constructions in Chapters 4 and 6 (the Conversational Construction and the Narrative Construction, respectively), however it figures prominently in the representation of obviation in Chapter 9. In Section 3.2, 1 present the primary reasons for the use of Construction Grammar for syntactic representations. Section 3.3 introduces the basic principles and mechanisms of Mental Spaces theory needed to talk about discourse structures, with the illustration of an introduction to a Potawatomi narrative. In Section 3.4, 1 argue for an elaborated representation of ground in Mental Spaces theory. This is useful for distinguishing basic types of illocutionary force such as statements and wh- questions, but becomes important for distinguishing various types of information in narrative (discussed in Chapter 7). " Recent work in Mental Spaces theory is paving the way for representations of constructions as blends, and so it might have been possible, with a little creativity, to construct the argument here using just the theory of Mental Spaces (in fact, in Chapter 4, 1 have represented subordinate clauses as embedded spaces). However, for detailed syntactic descriptions, such as is required for obviation, I have found it more practical to use Construction Grammar representations. The theories are generally compatible, however, and relatively easy to integrate. Information about mental space networks can, for example, be indexed in the external semantics of constructions, as I have done in Chapter 9. 33 3.2 Motivations for using a Construction Grammar framework Construction Grammar, as described by Fillmore and Kay (1993; 1999), and elaborated by Goldberg (1995), is a unificational theory of syntax that takes grammatical constructions (pairings of syntactic form with semantic meaning) as the central grammatical phenomenon to be explained. The motivations for using Construction Grammar here are outlined below. 'Non-core' grammar. The first is a theoretical commitment to take into account all of the conventional constructions that sanction sentences in a language as well as those that are less conventional or less common; not just what we might arbitrarily define as 'core' grammar. Well-known examples of 'non-core' grammar that have been addressed with this theory include the 'What is X doing Y?' (WXDY) construction analyzed by Fillmore and Kay (1999) ('What is this fly doing in my soup?), or the caused-motion construction discussed by Goldberg (1995) ('He sneezed the napkin off the table.'). While I will not be attempting to account for equivalent types of expressions in Potawatomi, I will, in the spirit of this theoretical commitment, try to account for the grammar found in discourse genres not traditionally addressed by syntactic theory, such as the morphosyntax of narrative discourse. The lexicon-syntax continuum. Secondly, Construction Grammar assumes that there is no strict separation between syntax and the lexicon. According to Goldberg, "Lexical constructions and syntactic constructions differ in internal complexity. . .but [they] are essentially the same type of declaratively represented data structure: both pair form with meaning." (p. 7) The discussions in Chapters 7 and 10 are a good argument for 34 extending this continuum to include discourse, since constructional forms can map onto discourse functions. Constructional Polysemy. Most lexical items exhibit polysemy, that is, they have sets of related meanings, some of which are presumed to be more basic, or central, than others. Likewise, studies of particular constructions have shown that they typically occur in networks of related senses, generally with a central sense extended to other senses.^ In this study, I argue for the existence of several constructions that each has multiple related senses in syntax and discourse. Because lexical items and constructions are presumed to have the same type of structure — that is, they are form-meaning pairings, this similarity in behavior is expected. Construction Grammar has a rather large set of representational conventions. The details of these conventions are not particularly germane to this discussion. The idea of constructions will be introduced in Chapters 4 and 6 in the discussion of the Conversational and Narrative Constructions. The theoretical mechanism of representation is not needed until Chapter 9, where it is introduced, along with a means of abbreviated representation. 3.3 Introduction to Mental Spaces theory The theory of Mental Spaces (Fauconnier, 1985; 1997) was developed to account for how we use language to construct and process meaning. According to the theory. ^ For example, with the caused motion construction, the central sense is successful transfer of a patient from an agent to a recipient, as in 'I gave Bill a cake. ' 'I baked Bill a cake' would be an extended sense where the agent intends to cause the recipient to receive the patient (Goldberg, 1995, p. 40). 35 when we engage in any kind of discourse, we partition information into mental spaces, which are "constructs distinct from linguistic structures, but built up in any discourse according to guidelines provided by the linguistic expressions" (Fauconnier, 1985, p. 16). Grammatical expressions such as adverbial clauses or conditional clauses, as well as aspects of grammar such as tense and mood, provide cues which allow speakers to create and navigate mental space structures, and signal listeners to do the same. The grammar and lexicon of a language are therefore used to establish and populate these mental spaces and track relationships between them. For example, various expressions such as in 1994..., Joe thinks..., if I win the lottery..., once upon a time..., set up spaces in which information is predicated, and considered valid. In the sentence 'In 1994, my daughter was two years old. ' The phrase in 1994 prompts the creation of a past space, in which the information 'my daughter is two years old' is valid (she would of course be much older today). The mental space structure for this sentence would look like the diagram in (1): (1) 'In 1994, my daughter was two years old. ' 'REALITY' SPACE (2003, my daughter is nine years old) PAST SPACE (1994, my daughter is two years old) 36 The spaces created during discourse are much more complex than this simple example. Many spaces are organized into a hierarchical network, beginning with an initial "reality" space, shown in (2) as Space R. New spaces (past spaces, future spaces, spaces for a narrative, etc.) are then set up subordinate to this space: (2) HIERARCHY OF SPACES IN A NETWORK Space R I use the term "reality" space with quotes to emphasize that this space does not represent a description of the real world, but rather speaker's mental representation (cognitive construal) of it. Since discourse context, particularly the roles of speaker and hearer, features prominently in this analysis, it is important to establish the "reality" space from the outset. This space is not always explicitly given in mental space representations. For example, Fauconnier begins space configurations with 'Space M,' sometimes defined as speaker reality (1985, p. 24). Cutrer apparently uses 'Space M' when de-emphasizing the context of a sentence, as with her illustrations of how BASE, V-POINT, FOCUS and EVENT work; later examples begin with 'Space R', speaker reality (1994, p. 104). 37 Besides the arguments which will be presented here, there is other evidence that every space configuration begins with a space which represents the "reality" of the speaker. Langacker (1991) has argued that every expression is grounded, although there is a cline with respect to the degree to which the ground is onstage and profiled. In addition, Liddell (1995), based on his work on ASL, has shown the necessity of setting up a 'real' space, a mental construct of the physical environment where people and objects physically present can be indexed. This real space is distinguished from 'surrogate' and 'token' spaces, which house the loci set up to reference people and things not present in the physical environment. A network of spaces has several features. At any point in a discourse, one of the spaces in the network is the BASE, one is the VIEWPOINT (or V-POINT) and one is the FOCUS. The feature BASE represents a deictic center of a conceptualizing self, and identifies the starting point for the discouse. In the default case, the BASE space is the here and now of speaker "reality", but may shift during discourse to represent another conceptualizer. The feature V-POINT identifies the space from which other spaces are currently being accessed and structured. According to Cutrer (1994), V-POINT stands for a bundle of deictic dimensions: In the strongest version, it represents the V-POINT of a conceptualizing self, with a full set of deictic dimensions. However V-POINT can also be more abstract, with a limited set of dimensions, in which case it corresponds to something like Langacker' s notion of 'vantage point' (Langacker, 1991). The third feature, FOCUS indicates which space is most active, the one that is currently being 38 structured with information.'* To these, Cutrer also adds EVENT, "the temporal space in which the event encoded in the verb takes place" (Cutrer, 1994, p. 72).^ 3.4 Illustration of the theoretical mechanism I will illustrate the basic operation of Mental Space in discourse by using the illustration of the beginning of a Potawatomi narrative. How Rabbit Got a Short Tail (MD102694). The lines of the narrative to be discussed are as follows: (3) HOW RABBIT GOT A SHORT TAIL 1 O, neko ngi-babzedwak neshnabek I used to listen to the people telling e-yayajmowat eyayengajmowat . stones ; something they laughed about. 2 [Iw je] ni wabozoyen ngodek e-gi-yajmawat . Once they told abou t Rabbit. 3 O, bnewi neko o wabozo gi-gnewanwe . Oh, at one time Rabbit had a long tail . 4 Gi-gnewanwedek kedwik . He must have had a long tail , they say . 5 Iw je i wech-shkwanwat ngom ga-zhewebzet . That's why he has a short tail today, because of what happened to him . 6 Jigbyek ibe e-pa-zhyat . He went around there by the water. (MD 102694) Line 1 of the narrative begins with the narrator describing an activity in the past, listening to people telling stories. This sets up the BASE space in the here and now (what I will refer to as speaker "Reality"). The particle neko 'used to' plus the past tense gi- on ngi-babzedwak 'I listened to them' opens a past space embedded in the BASE space. "* FOCUS was incorporated into Mental Spaces theory based on the work of Dinsmore (1991). ^ The feature EVENT is primarily needed to represent tense. Although I will use it in diagrams, I will not discuss it in more detail here, since tense is peripheral to this analysis. 39 This subordination of the Past Space to the present BASE space is represented with a connecting line between the two spaces. When the past space is opened, V-POINT remains with BASE in the "Reality" Space, as indicated by the use of past tense. In other words, the information predicated in this space is about past events and entities, and not about the speaker's present. FOCUS thus shifts to the Past Space, indicating that this is the space currently being structured by new information. (4) FOCUS SHIFTS TO PAST SPACE ...neko ngi-babzedwak neshnabek e-yaya jmowat... 7 used to listen to them telling stories ' 'REALITY' SPACE (the here and now) BASE V-POINT FOCUS PAST SPACE ngi-babzedwak 'I used to listen to them' In line 2, e-gi-yajmawat 'they told about him' opens a Narrative Space subordinate to the past space. The Narrative Space and any spaces subordinate to it are separated from the rest of the network in a narrative domain. This domain is then subordinate to spaces predicated in the "Reality" domain. FOCUS now shifts to the narrative space, however the use of the past tense signals that BASE and V-POINT remain in the "Reality" Space. We also learn that this space is populated by an entity 40 Wabozo 'Rabbit' (represented by w' in the Narrative Space) This Rabbit is understood to be a mythic character; either a role, or possibly a prototypical instance of rabbits whose traits are inherited by all modern rabbits. Modem rabbits are represented in the "Reality" Space as w. (The line connecting w and w ' is explained below.) (5) REFERENT W ESTABLISHED IN THE NARRATIVE DOMAIN [ Iw je] ni wabozoyen ngodek e-gi-ya jmawat . 'Once they told about rabbit ' BASE V-POINT 'REALITY' SPACE w = wabozo 'rabbit' (modem day rabbits) PAST SPACE Reality " Domain Narrative Domain NARRATIVE SPACE w' = rabbits of long ago FOCUS Lines 3 and 4 now structure the narrative space, which is in FOCUS. We learn that bnewi neko 'it used to be long ago' gi-gnewanwe 'he (Rabbit) had a short tail', so this information is added in the representation to the narrative space (shown in (6)) 41 (6) FOCUS SHIFTS TO NARRATIVE SPACE O, bnewi neko o wabozo gi-gnewanwe . 'At one time, rabbit had a long tail. ' 'REALITY' SPACE w = wabozo 'rabbit' BASE V-POINT PAST SPACE 'Reality " Domain Narrative Domain NARRATIVE SPACE have a long tail (w) FOCUS In line 5, the third person pronominal referent 'he' in the participle wech-shkwanwat 'why he has a short tail' sets up a counterpart to Wabozo (w') in the "Reality" domain. We represent the pragmatic relationship between the two referents w and w' with a connector (line) between their referents in the two different spaces. This Rabbit is also a role, but instead of having a long tail, he has a short tail. We add this information to the 'reality' space, since this is information about modern rabbits. We now come to a classic problem of reference that is easily solved in Mental Spaces theory. The problem is the non-contradiction in a sentence like 'In that painting, 42 the girl with the blue eyes has green eyes. ' Without the phrase 'in that painting' t\\Q rest of the sentence is contradictory. Fauconnier and others have noticed that representations (such as paintings, photographs, etc.) set up pragmatic relationships between the representation and the model, where the representation and model are counterparts. In the following diagram, the phrase 'in that painting ' sets up a representation space subordinate to the "reality" space. The blue-eyed girl (a) is set up as an entity in the "reality" space and the green-eyed girl (a) is an entity in the representation space. The line connecting them indicates that a and a' are counterparts: (7) 'In that painting, the girl with the blue eyes has green eyes. ' 'REALITY' SPACE where a = girl and a has blue eyes REPRESENTATION SPACE (the girl has green eyes) An entity in one space can then be referred to by its counterpart in another space, so that the girl with the blue eyes can refer to the entity in the representation space, meanwhile, information predicated about one or the other entity can be true within its own context.^ The same has been shown to be the case in a wide variety of contexts See Nunberg (1978; 1979), Jackendoff (1975) and Fauconnier (1997). 43 including beliefs, as in 'George believes that the girl with blue eyes has green eyes, ' and narratives, as in 'In that story, the girl with the blue eyes has green eyes. ' Returning to our narrative, we are faced with the potential contradiction in lines 4 and 5 that rabbits have long tails and rabbits have short tails. What allows us to keep this non-contradictory is the establishment of a narrative space where the information 'rabbits have long tails' is valid. This narrative space is already available, set up in previous sentences. To this pre-existing narrative space, we set up w' for long-tailed rabbits, and link this to its counterpart w in the "reality" space which represents short-tailed rabbits, as shown in (8). '^ (FOCUS shifts back to the 'reality' space where we add the information that rabbits have short tails.) ^ There is also a counterpart to w in the Past Space. This is not represented in the diagram merely for the sake of simplicity of representation. In general, 1 will only note counterparts in diagrams for spaces that are currently being discussed. 44 (8) COUNTERPART TO W SET UP IN THE REALITY DOMAIN Gi-gnewanwedek kedwik. Iw je i wech-shkwanwat ngom ga-zhewebzet . 'He must have had a long tail, they say. That 's why he has a short tail today, because of what happened. ' 'REALITY' SPACE BASE V-POINT FOCUS PAST SPACE "Reality " Domain Narrative Domain NARRATIVE SPACE gi-gnewanwe (w) Line 6 begins the narrative proper. From this point, most of the information structures spaces in the Narrative Domain. BASE and V-POINT remain in the Reality Domain, and FOCUS shifts to the Narrative Domain, as shown in (9). This is the basic arrangement for the activity of 'narration'. In Chapter 6, 1 show how this configuration changes to accommodate the representation of a narrative-internal viewpoint, such as the representation of a character's perspective. 45 (9) THE BASIC NARRATIVE CONFIGURATION Jigbyek ibe e-pa-zhyat. 'He went around there by the water. ' 'REALITY' SPACE w = wabozo 'rabbit' BASE V-POINT PAST SPACE 'Reality " Domain Narrative Domain NARRATIVE SPACE have a long tail (w) FOCUS 3.5 An elaborated representation of ground In the this section, I describe in more detail what is meant by the "reality" space, and argue for an elaborated representation of ground (that is, the representation of the "Reality" Domain) in Mental Spaces theory. This representation will become important in the discussion of Potawatomi Narrative in Chapter 7. 3.5.1 The "reality" space The simplest space configuration consists of a single space; the "reality" space of the speaker. This space functions as the BASE, and is the locus for V-POINT and FOCUS as shown in (10). This default configuration serves as a starting point for any 46 discourse; thus every communicative act is ultimately grounded in the deictic center of the speaker. (10) SIMPLEST SPACE CONFIGURATION While discourses commonly build up a large network of spaces, this single-space configuration can be approximated by a simple conversation in and about the here and now. Consider the following dyadic exchange. The conversation takes place in the kitchen belonging to A and B. The jar of mayonnaise has recently been purchased, and A wonders whether it has been put on the shelf or refrigerated. Space R: BASE V-POINT FOCUS (1 1) A: Where is the mayonnaise? B: In the fridge. If I am speaker A, the configuration for this exchange can be represented by Space R, my "reality" space, which is minimally populated by myself (a), a conversation partner (b), and the mayonnaise (m) and fridge (f). In this case, (a), (b) , (m) and (f) exist in the proximate space. ^ The mayonnaise and fridge, as definite descriptions, are both ^ That is, (a) and (b) are proximate for the purpose of face-to-face conversation, and (m) is proximate for the purpose of (a)'s easily fetching it. Note that this is a cooperative scenario; if the exchange occurred at a 47 present in Space R, supplied by the context which includes a frame for the activity, 'sandwich-making,' and the physical environment of the kitchen (this frame is not otherwise represented in the diagram). As conceptualizing individuals, (a) and (b) supply a potential V-POINT (represented by "@"), each of which is available as a BASE space. By default. Space R represents the BASE associated with the role of speaker. The BASE space for (b) is represented by a Space H (for "Hearer") subordinate to Space R. (a) and (b) are assigned to the roles of either Speaker or Hearer, depending on the point in the exchange.^ These roles are supplied by the discourse frame 'dyadic conversation.' ^^ picnic, (b)'s reply would flout the maxim of relevance, since the refrigerator is not proximate, meaning that 'we left it at home. ' ' Dancygier and Sweetser (1996) includes a representation of the discourse context, includmg Speaker and Hearer (labeled as individuals, though rather than roles) in their discussion of metalinguistic spaces. This is the only other work within the Mental Spaces theory literature (that I am aware of) to make ground explicit in a configuration. ^° As another example, the discourse frame 'lecture' would supply a lecturer, an audience, expectations about venues, possible subject matter, etc. Unlike roles supplied by the content of the discourse, such as 'the president' in the sentence 'the president changes every four years,' discourse roles are non-explicit, and backgrounded. 48 (12) MODEL FOR DYADIC CONVERSATION DOMAIN OF SPEAKER REALITY Space R: FRAME: dyadic conversation ROLES: Speaker (S) Hearer (H) DOMAIN OF HEARER REALITY Space H 3.5.2 The profiling of discourse participants In the model for a prototypical diadic conversation, one participant is always profiled. For example, if (b) is the conceptualizer in the conversation in (1 1), when (b) makes the statement 'in the fridge', (b) is the speaker and profiled participant, as in (13). I represent this profiling by the use of a feature FOCUS CONTEXT (which will be explained below). (13) SPEAKER IS CONCEPTUALIZER AND PROFILED PARTICIPANT 49 DOMAIN OF SPEAKER REALITY Space R: BASE V-POINT FOCUS CONTENT FOCUS CONTEXT FRAME: dyadic conversation ROLES: Speaker (S) Hearer (H) DOMAIN OF HEARER REAUTY Space H (b) is also the profiled participant from (a)'s point of view as hearer. In (a)'s mental space network, this is represented by FOCUS CONTEXT moVmg to Space S (for the "Speaker"), as in (14). (14) HEARER IS CONCEPTUALIZER, SPEAKER IS PROFILED PARTICIPANT DOMAIN OF HEARER REAUTY Space R: BASE V-POINT FOCUS CONTENT Space S FRAME: dyadic conversation ROLES: Speaker (S) Hearer (H) DOMAIN OF SPEAKER REALITY FOCUS CONTEXT 50 Certain types of illucutions, such as Wh-questions, foreground the hearer's role as a conceptuaUzer. In (a)'s question 'Where is the mayonnaise?', the hearer (b) is profiled as a conceptualizer who possesses potentially unique knowledge, thoughts, and beliefs. The question word 'where' implies that the hearer has knowledge that the speaker does not possess; that their representations of reality are different on this point. This is represented in (15) below. (15) SPEAKER IS CONCEPTUALIZER, HEARER IS PROFILED PARTICIPANT DOMAIN OF SPEAKER REAUTY Space R BASE V-POINT FOCUS CONTENT Space H FRAME: dyadic conversation ROLES: Speaker (S) Hearer (H) DOMAIN OF HEARER REALITY FOCUS CONTEXT Of the existing theoretical features, the most likely candidate to represent this profiling is FOCUS, since focus has to do with foregrounding components of the discourse structure. In this respect, it is similar to Langackef s profiling, which gives special prominence to a part of a semantic structure, but on the level of discourse rather than word or sentence level semantics. 51 I therefore propose splitting FOCUS into two dimensions: a content dimension, and a context dimension. The content dimension represents what we normally think of as FOCUS, that is the space currently being structured. In Dinsm ore's terms the space in FOCUS is "[t]he space that a discourse sentence as a whole is intended to say something about, that is, the space into which the sentence is contextualized" (1991, p. 122).^^ The context dimension, on the other hand, is relevant when a discourse participant, for one reason or another, is brought into the foreground and thus commands our attention. FOCUS context therefore involves the highlighting of discourse participants. This representation of discourse participant profiling will be taken up again in the discussion of narrative in Chapter 7. ^' In the following discussion, where 1 use the tenn FOCUS alone, I am referring to FOCUS CONTENT. 52 Cutrer, Michelle. 1994. Time and Tense in Narratives and Everyday Language, University of California: Ph.D. dissertation. Dancygier, Barbara; and Eve Sweetser. 1996. Conditionals, Distancing, and Alternative Spaces. Conceptual Structure, Discourse, and Language, ed. by Adele Goldberg, 83-99. Stanford: CSLI Publications. Dinsmore, J. 1991. Partitioned Representations. Dordrecht: Kluwer Academic Publishers. Fauconnier, Gilles. 1985. Mental Spaces. New York: Cambridge University Press. — . 1997. Mappings in Thought and Language. Cambridge: Cambridge University Press. Fillmore, Charles and Paul Kay. 1993. Construction Grammar, vol. Unpublished ms. Berkeley, CA — . 1999. Grammatical constructions and linguistic generalizations: The What's X doing Y? Construction. Language, 75.1-33. Goldberg, A. 1995. Constructions: A Construction Grammar Approach to Argument Structure: Cognitive Theory of Language and Culture. Chicago: University of Chicago Press. Jackendoff, Ray. 1975. On belief contexts. Linguistic Inquiry, 6. Lakoff, G. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. Langacker, R. 1991. Concept, Image, and Symbol: The Cognitive Basis of Grammar. vol. 1: Cognitive Linguistics Research. Berlin: Mouton de Gruyter. Liddell, Scott. 1995. Real, Surrogate, and Token Space: Grammatical Consequences in ASL. Language, Gesture, and Space, ed. by J. Reilly, 19-41. New Jersey: Laurence Erlbaum Assoc. Nunberg, Geoffrey. 1978. The pragmatics of reference. Bloomington, Ind.: Indiana University Linguistics Club. — . 1979. The non-uniqueness of semantic solutions: Polysemy. Linguistics and Philosophy, 3. 53 4 Independents and Conjuncts in Everyday Discourse 4.1 Introduction This chapter describes the distribution of independent and conjunct verbs within the context of everyday discourse. As a general statement, the independent order is found in main clauses, and the conjunct in subordinate clauses. While this is statement is sufficient to account for the independent order, there are a number of aspects to the use of the conjunct which will require some refinement of this statement, including its co- occurrence with a factive-like preverb e-, and its use in certain main clause contexts. Establishing the basic uses of the independent and conjunct, as well as the preverb e- will be important for contrasting their use in narrative discourse (examined in Chapter 6). 4.2 Main clause independents and subordinate clause conjuncts In conversational discourse, the independent is the form for main clause verbs as shown in (1) - (3) below. Independent verbs are underlined: (1) Mani wi-gishnenan niw dabyanen. mani wi- gishEnEn -a -En niw Odabyan -En Mary PUT- buy.s.o\TA -DIR -OBV.I that . OBV car -OBV Mary will buy the car. (POEX00039) (2) Mikjewimget ne? itiikEjewi -mEgEd nE work\AI -AUG. O.I Q Does it work? (POEX00045) 53 (3) Mani wgi-gzibyenan mine mani wE- gi- gEzibyen -a -En itiinE Mary 3- PST- wash.s.t\TI -OBJ -OBJ. I and wgi-begwabke ' anen niw nagnen. wE- gi- begwabEkE' -a -En -En niw nagEn -En 3- PST- dry.s.t\TI -OBJ -OBJ -PL. OBJ. I that.OBV dish -PL Mary washed and dried the dishes. (POEX00146) Conjunct verbs are used in subordinate clauses. Examples are given below of complement clauses (4) - (5) and adverbial clauses (6) - (7). Conjunct verbs are underlined: (4) Ndenendan Mani e-wi-gishnenat nEd- Enend -a -En mani e- wi- gishEnEn -a -Ed 1- think.thus .of .s .t\TI -OBJ -3/0.1 Mary FCT- FUT- buy.s.o\TA -DIR -3.C niw wdabyanen. niw Odabyan -En that.OBV car -OBV / think that Mary will buy the car. (POEX00040) (5) Ni pi je e j e-bmosewat? ni pi jE CH.EjE- bEmOse -wad where in . a . certain . direction- walk\AI -35. C Where are they walking? (POEX00266) (6) Zagech zhyayen, gizho'on. zagEj Ezhya/e -yEn gizho'o -En outside go\AI -2.C dress .warmly\AI -2 . IMP If you go outside, dress warmly. (POEX00019) (7) E-mnadenj eget , mno-ye e - mEnadenEjege -Ed mEnO- EyE FCT- be . respectful\AI -3.C good- be . in . a . place\AI . I Because she is respectful, she lives well. (POEXOOOll) 54 4.3 Conjuncts that take the e- preverb A verb in the conjunct form is frequently preceded by the preverb e-. It is unclear exactly how this morpheme should be translated. Hockett noted in his work on Potawatomi in the 1940's that the preverb e- is a mark of the storytelling style, glossing it as a 'narrative' preverb: ^ "First-position Preverbs. ?e, with conjunct mode only, narrative: ?e ki mpot he died. Translation usually cannot show the force of this preverb; it is the mark of a certain style, namely that of story- telling and the like, in contrast to statements made about what has happened, in reality, to the speaker." (Hockett, 1948b, p. 139) There is also a tradition of calling e- an aorist, going back to Bloomfi eld's use of the term for Fox (Bloomfi eld, 1927). He seems to have used it to refer to its function in traditional narrative where it can be glossed as a past tense: "The changed conjunct of stems containing a particle eeh (this is the changed form; the simple form does not occur) is common in C[ree]: eeh-takohteet "when he arrived." It occurs occasionally in 0|jibwa]; in F[ox] this form serves also for nonsubordinate statements in hearsay narrative: eeh-pyaaci "when he came; he came (it is said)." (Bloomfield, 1946, p. 101) Goddard (1990) also uses 'aorist' for Fox, however he treats the preverb plus conjunct as an unchanged conjunct form. ^ The historical provenence of the preverb e- is unclear. It is perhaps the changed form of a preverb (short vowel) a- which is only attested in the related language Ottawa, of which Bloomfield says "[it] is used with conjunct verbs only; it denotes place or person" (1958, p. 62). Two examples can be found in the text, both of which are locative m function: a-wOTac/A^J 'where he sat' (1958, p.l78) and a-fem/'-woog^eg- 'train station' (1958, p. 62) (literally, "where the train stops" (Rhodes, 1985, p.l)) In younger speakers of Ottawa, e- is taking over as an invariant form of initial change (Costa, 1996; Rhodes, 1985), this may be happening for some speakers of Potawatomi as well, but for the speakers cited here initial change is still maintained. 55 However, the prevalence of e- in conversation requires us to conclude that its semantics is more complex than being simply an indicator of the narrative discourse mode. In embedded sentence complement clauses, e- indicates that the proposition expressed by the dependent clause verb is either presupposed to be true as in (8) and (9), or that it is probable as in (10): (8) Ngi-wabma Mani e-gishnenat nE- gi- wabEm -a mani e - gishEnEn -ad 1- PST- see.s.o\TA -DIR.I Mary FCT- buy.s.o\TA -3/0. C niw wdabyanen. niw Odabyan -En that.OBV car -OBV I saw Mary buy the car. (POEX00068) (9) Ngekendan Mani e-wi-gishnenat nE- gEkend -a -En mani e - wi- gishEnEn -ad 1- know.s.t\TI -OBJ -1/0. I Mary FCT- FUT- buy.s.o\TA -3/0. C niw wdabyanen. niw Odabyan -En that.OBV car -OBV I know that Mary will buy the car. (POEX00086) (10) Ndenendan Mani e-wi-gishnenat nEd- Enend -a -En mani e- wi- gishEnEn -a -Ed 1- think.thus .of .s .t\TI -OBJ -3/0.1 Mary FCT- FUT- buy.s.o\TA -DIR -3.C niw wdabyanen. niw Odabyan -En that.OBV car -OBV / think that Mary will buy the car. (POEX00040) Possibility as well as obligation are indicated by the use of the sequence of preverbs da-je- (11)-(12): 56 (11) Ndenendan Mani da- j e-gishnenat nEd- Enend -a -En mani da-jE- gishEnEn -a -Ed 1- thmk.thus .of .s .t\TI -OBJ -1/0. 1 Mary MOD- buy.s.o\TA -DIR -3.C niw wdabyanen. niw Odabyan -En that.OBV car -OBV / think that Mary might buy the car. (POEX00049) (12) Mani wgi-mikwendan da- j e-gishnenat mani wE- gi- mikwenEd -a -En da-jE- gishEnEn -ad Mary 3- PST- remember . s . t\TI -OBJ -OBV. I MOD- buy.s.o\TA -3/0. C niw wdabyanen. niw Odabyan -En that.OBV car -OBV Mary remembered that she ought to buy the car. (POEX00050) In adverbial clauses, the use of e- is restricted to those that are non-hypothetical. Examples of non-hypothetical adverbial clauses are given below in (13) through (18). The adverbial clause verb is underlined: Reason clause: ^ (13) E-mnadenj eget , mno-ye e - mEnadenEjege -Ed mEnO- EyE FCT- be . respectful\AI -3.C good- be . in . a . place\AI . I Because she is respectful, she Uves well. (POEXOOOl 1) ■ Categories of adverbial clauses are based on the terminology given in Thompson and Longacre (1985). 57 Purpose clause:^ (14) Odanek nwi-zhya wisnewen e-wi-gishnedoyan . odan -Eg nE- wi- Ezhya/e wisEnEwEn e - wi- gishEnEd -o -yan town -LOG 1- FUT- go\AI . I food FCT- FUT- buy.s.t\TI -OBJ -l.C I am going to town in order to buy food. (POEX00015) Durative clause: (15} odanek e-gi-bme-yeyan, Wayne odan -Eg e- gi- bEmE- EyE -yan Wayne town -LOG FCT- PST- during- be . in . a . place\AI -l.C Wayne gi-binchege . gi- binEchEge PST- clean. things\AI . I While I was in town, Wayne cleaned. (POEX00036) Iterative clause: (16) E-gish-wisnet , neko mbe . e - gizh- wisEn -Ed nEko nEba/e FCT- finish- eat\AI -3.C used. to sleep\AI.I Whenever she finished eating she used to sleep. (POEX00015) Universal clause: (17) E-gmeyak, zhoshkwa. e - gEmEya -Eg zhoshEkwa FCT- rain\II -O.C be . slipperyMl . I Whenever it rains, it is slippery. (POEX00015) ^ Purpose clauses are in a sense hypothetical, since they always occur in the future with respect to the main clause. However, because of their semantic similarity, reason and purpose adverbial clause types are formed the same way in many of the world's languages (Thompson and Longacre, 1985, p. 185). Linguistic motivation for the use of e- m purpose clauses may thus be in conformance with this observed tendency. 58 Time-cause clause: (18) E-gkenmek ga-nshkadzet , e - gEkenEm -EgO CH.gi- nEshkadEzE -Ed FCT- know.s.o\TA -1/3. C PST- be . angry\AI -3.C ngi-ne-maj i . nE- gi- nE- maji 1- PST- start. to- leave\AI.I When I realized he was angry, I left. (POEX00038) The preverb e- is not used in hypothetical clauses, as shown in (19) through (21): Hypothetical conditional clauses: (19) Zagech zhyayen, gizho'on. zagEj Ezhya/e -yEn gizho'o -En outside go\AI -2.C dress .warmlyXAI -2 . IMP If you go outside, dress warmly. (POEX00019) (20) Gishpen bonimgek, nwi-we-zhoshk ' o . gishpEn boni -mEgEg nE- wi- wE- zhoshEk ' o if snow\II -AUG.O.C 1- PUT- go . and- go . sledding\AI . I If it's snowing, III go sledding. (POEX00021) Counterfactual conditional clause: (21) Gishpen bonimgek, nda-zhoshk ' o . gishpEn boni -mEgEg nE- da- zhoshEk ' o if snowMl -AUG.O.C 1- MOD- go . sledding\AI . I If it were snowing, I would be sledding. (POEX00023) The use of e- in non-hypothetical adverbial clauses produces a contrast between the concessive conditional (22), glossed 'even if and the concessive (23), which presupposes 'she is young' glossed with 'although': 59 (22) Anwe zhe penojewet, mbwaka. anwE EzhE EpEnojew -Ed nEbwaka/e although EMPH be . a . child\AI -3.C be .wise\AI . I Even if she is young, she is nevertheless wise. (POEX00025) (23) Anwe zhe e-penoj ewet , mbwaka. anwE EzhE e - EpEnojew -Ed nEbwaka/e although EMPH FCT- be . a . child\AI -3.C be.wise\AI.I Although she is young, she is nevertheless wise. (POEX00026) 'Before' clauses take the particle bwamshe 'before' and do not take e- as in (24) and (25): (24) Odanek bwamshe zhyayan, nge-wjanda. odan -Eg bwamEshE Ezhya/e -yan nE- gE- Ojanda town -LOG before go\AI -l.C 1- PUT- cook\AI . I Before I go to town, I'll cook. (POEX00033) (25) Ngi-wjanda bwamshe majiyan. nE- gi- Ojanda/e bwamEshE maji -yan 1 PST cook\AI.I before leave -l.C / cooked before I left. (POEX00229) James (1983) for Moose Cree suggests that the absence of e- in 'before' clauses is due to the fact that they are always in the future with respect to their main clauses, and from that perspective can be considered hypothetical. More generally though, e- is not used in any temporal clause that expresses futurity, as shown by (26) as compared with (27): (26) odanek zhyayan, wisnewen nda-gishnedon. odan -Eg Ezhya/e -yan wisEnEwEn nE- da- gishEnEd -o -n town -LOG go\AI -l.C food 1- MOD- buy.s.t\TI -OBJ -1/0. I When I go to town, lean buy food. (POEX00035) 60 (27) Odanek e-gi-zhyayan, wisnewen ngi-gishnedon . odan -Eg e- gi- Ezhya/e -yan wisEnEwEn nE- gi- gishEnEd -o -n town -LOG FCT-PST- go\AI -l.C food 1- PST- buy.s.t\TI -OBJ -1/0. I When I went to town, I bought food. (POEX00274) Likewise, 'after' clauses in the future do not take e- as in (28): (28) Bama zhe gish-ggwadman node mkeznen bama zh E gizh- gOgwad -man nodE mEkEzEn -En later EMPH finish- sew.s.t.\TI -1/0. C these. INAN moccasin -PL nwi-mba . nE- wi- nEba/e 1- FUT- sleep\AI.I After I finish sewing these moccasins, I'll go to bed. (.TTNB3p53n2) However, 'after' clauses in the past occur with initial change, which is generally found in factive-like contexts where the proposition in the clause is presupposed. In (29) it is registered in the preverb ga-, which is the changed form of past tense gi-: (29) Ga-mbayan, gi-wep-boni . CH . gi - nEba/e -yan gi- web- boni PST- sleep\AI -l.C PST- start. to- snowMl.I After I slept, it started to snow. (POEX00275) 4.4 The distribution of conjuncts in main clauses Besides the subordinate clause use of the conjunct as described above, there are a few contexts where the conjunct can be used in a main clause, often with an accompanying particle, as illustrated in (30) with the particle bedo 'wish that' (conjunct underlined): 61 (30) O, bedo wi na bkenageyan! o bedo wi na bEkEnage -yan oh wish. that EMPH EMPH win\AI -l.C Oh, I wish I would win! (POEX00261) Because these contexts pose a problem for a simple distributional statement of the conjunct as a subordinate clause verb form, the traditional means of handling them has been to define the particles as subordinators."* This solution is more satisfying for the few particles which always require the presence of a conjunct. However, for many particles, the presence of a conjunct is optional; moreover, the conjunct can also occur in a main clause without a particle. Clearly, in order to be able to explain these sentences, an explanation that does not rely on an overt subordinating particle is needed. In this section, I will show that rather than being simply idiosyncratic, the use of the conjunct in these contexts is well-motivated in that the apparently dissimilar main clause contexts have a common semantics involving speaker subjectivity. Moreover, this shared semantics motivates calling these contexts subordinative, even in the absence of an overt subordinator. 4.4.1 Adverbial particles that can take a main clause conjunct Many adverbial particles commonly co-occur with a main clause conjunct, but do not require its use. The particles that fall into this 'optional use' category, all have modal semantics, encoding the speaker's attitude towards the propositi onal content of the utterance. Examples of these particles are given below in (3 1) - (38), taken from "* Bloomfield, for Eastern Ojibwa, calls them 'predicative particles' (1958, p. 141). 62 elicitations and quoted speech in narrative texts (which behaves like everyday conversation with respect to the use of verbal paradigms). Anake 'maybe'. The most common use ofanake is as a disjunctive, in which case it is used with a main clause independent, as in (3 1). (31) Nin anake gin gda-kwabmamen penojeyek. nin anake gin gE-da- kEwabEiti -a -EmEn EpEnoje#y -Eg I.EMPH or 2.EMPH 2- MOD- watch . over . s . o\TA-DIR-12 . I child -PL You or I should watch the kids. (POEX00208) However, as a subordinating particle, it is best translated as 'maybe', as in (32). In this example, the speaker indicates a mental stance towards the addressee's behavior, without specifying exactly what that is. This indirect tactic leaves it to the addressee to work out the mild criticism: (32) Gwi-gwdemoj ge ne? Anake (zhe) bama gmeyamgek. gE- wi- gOdEmojEge nE anake zhE bama gEmEya -mEgEg 2- FUT- fish\AI.3.I Q maybe EMPH wait rain\II -AUG.O.C Are you going fishing [when the weather is fine]? Maybe you should wait until it rains. (POEX00258) Iw zhe anwe 'okay'. This particle phrase is commonly used on its own, as in response to the query, Nije ezh-bmadzeyen? 'How are you doing?' Here it is used to give an appraisal of someone's speaking ability: (33) Iw zhe anwe e-neshnabemot . iw zh E anwE e - EnEshEnabemO -d that.INAN EMPH all . right FCT- speak . Indian\AI -3.C 'He 's getting to talk Indian okay now. ' (POEX00272) Wete 'really'. The particle wete is generally used to indicate the speaker's attitude. Thus in (34), the Lazy Grasshopper tells the Busy Bee he doesn't care what the Bee thinks, and implies something like 'and I shouldn't, either' (compare 'I don 't care 63 what you think' with 'I don 't really care what you think'' which shows a similar discourse use of English 'really'): (34) Ngoji e-nme-se-gwakwaskso ' ot , "Wete ngOji e- nEitiE- sE - gwakwaskOsE ' o -d wetE somewhere FCT-in . the . process . of - EMPH - hop\AI -3.C really wi zhe na nin gbapnenmen, " wi zhE na nin gE- bapEnenEiti -En EMPH EMPH EMPH I . EMPH 2 -couldn ' t . care . less . f or . s . o . \TA -1/2.1 e-nat ni amon . e- En -ad niw amo -n FCT- say.to.s .o. \TA -3/3'. C that . OBV bee -OBV He hopped away someplace "I could care less what you think, " he said to the bee. (HOBN2t2.010) Wika 'finally'. Wika is used to express 'finally' in the sense of 'at long last', indicating either hope or expectation on the part of the speaker that an event would occur sooner than it did.^ Wika is commonly found with a main clause conjunct as in (35): (35) wika se na e-gi-ma j it . wika sE na e - gi- maji -d finally EMPH EMPH FCT- PST- leave\AI -3.C Finally, he left! (POEX00285) Negative particles. The conjunct is also found in with certain negative particles, such asjo mamda 'it is not possible' (36) - (37) andy'o wi zhe gego 'it doesn't matter' (38): ^ This particle contrasts with another particle gegpi which is also translated as 'finally ' but does not carry the same sense of hope or expectation. It is commonly found in narratives when a character turns to a new activity, as in Gegpi, e-gi-majit. 'Eventually, he left.' (POEX00286). (The use of the main clause conjunct here is a feature of narrative which is discussed m Chapter 6). 64 (36) I je o shebzhi neyap e-gi-zhyat iw jE ow mEshEbEzhi neyab e - gi- Ezhya/e -d and that. AN lion back FCT- PST- go.there\AI -3.C e-gi-widmowat niw wshkabewsen, e - gi- widEmEw -Ewad niw wEshkabewEs -En FCT- PST- tell.s.o\TA -3/3'. C that.OBV helper -OBV "Jo mawda e-wi-nsek," e-gi-nat. jo mamda e - wi- nEs -Eg e- gi- En -ad not possible FCT- FUT- kill.s.o.\TA -1/3. C FCT-PST- say . to . s . o\TA -3/3'. C So the lion went back and told the attendants "I couldn 't kill him. " (JS.4. 1.032) (37} Jo mamda zhode bidek da- j e-wdemayen . jo mamda zhodE bidEg da-jE- OdEma/e -yEn not possible here indoors MOD- smoke . tobacco\AI -2.C You can 't smoke in here. (JT. 03. 41. 006) (38) Jo wi zhe gego j agdewpegwzewat . jo wi zh E gego j agEdewpEgOzE -wad not EMPH EMPH something taste . burnt \AI -35. C It doesn 't matter if they (potatoes) taste burnt (JT.3. 35.01^ 4.4.2 Particles that require the use of a main clause conjunct There are a few particles that require the use of a main clause conjunct. These include bedo and begesh'^ 'wish that' and yedek 'it must be that', edgwen 'I wonder' and nmed se na 'I don't know' (with allegro forms nmej zhe na and nmej na). Examples are given below in (39) -(46): (39) Bedo (wi) na gmeyamgek. bedo wi na gEmEya -mEgEg wish. that EMPH EMPH rain\II -AUG.O.C I wish it would rain! (POEX00262) ^ Different speakers use one or the other particle. Begesh has cognates in Ojibwe and Ottawa. 65 (40) "0, begesh na ezhi gameyek o begEzh na ezhi gameyEg oh would. that EMPH over. there across . the . river gshketoyan e-byayan, " gEshEkEt -o -yan e - bya/e\AI -yan be.able.to.do.s.t\TI -OBJ -l.C FCT- come -l.C e-kedot... A, begesh na ibe zhyayan." e - EkEdO -d a begEzh na ibE Ezhya/e -yan FCT- say\AI -3.C ah would. that EMPH there go.there\AI -l.C "Oh, I wish I was able to get across over to there, " he said. . . Ah, I wish I could go over there. " (MD102694.007, 010) (41) " Iw se zhye yedek e-wi-byawat nmezodanek, " iw sE zh yE yedEg e - wi- bya/e -wad nE- mEzodan -Eg that.INAN EMPH EMPH must.be FCT- FUT- come\AI -35. C 1- parent -PL e-zhede ' at . e - EzhEde ' a -d FCT- think\AI -3.C "So now must be my parents will come, " he thought. (AS. 2. 3. 080) (42) "I je bzhe gagyaw yedek e-gi-mot." iw jE bzhE gagyaw yedEg e - gi- mEw -Ed and EMPH anyhow must.be FCT- PST- eat.s.o.\TA -2/3.C Well, must be you ate him anyway. (AS. 2. 1.029) (43) "A, iw zhe yedek e-wi-dkemozh ' ewat a iw zhE yedEg e- wi- dEkEmozhE' -Ewad ah that.INAN EMPH must.be FCT- FUT- take . s . o . across\TA -35/1. C gode," zhede ' e o wabozo. godE EzhEde ' a/e ow wabozo#y these. AN think\AI.3.I that. AN rabbit "Ah, must be they will take me across, " thinks the rabbit. (MD 102694.027) (44) I je o neshnabe e-nat, " Edgwen se na iw jE ow EnEshEnabe e- En -ad edEgwen sE na and that. AN person FCT-say . to . s . o\TA -3/3'. C I. wonder EMPH EMPH a- j e-gshke ' nan nsheke." a- EjE- gEshkE' -Enan nEshEke MOD- towards- be . able . to . do . s . t . to . s . o . \TA -1/2. C alone And the man told him, "I don't see how I'll be able to do that alone. "(JS. 4.5. 01 3) 66 (45) Nmet zhe na da- j e-bonimgek . nEmEd zh E na da-jE- boni -mEgEg I. don't. know EMPH EMPH MOD- snow\II -AUG.O.C 7 don 't know if it will snow. ' (POEX00273) (46) Iw je e-gi-majit, nmej na yedek iw jE e - gi- maji -d nEmEd sE na yedEk and FCT- PST- leave\AI -3.C I . don ' t . know ga-zhyagwen . CH.gi- Ezhya/e -gwen CH.PST- go.there\AI -DUB.3.C So he started off somewhere. (Literally: 'he started off, I don 't know where he went') (AS.2. 1.008) 4.4.3 Wh-question particles Wh-questions always take a main clause conjunct. Wh-questions are formed by the use of an initial question particle or particle cluster, and require the use of a main- clause conjunct, which has in addition initial change: (47) Ni je ezh-bmadzet ? ni jE CH.EzhE- bEmadEzE -Ed what thus- live\AI -3.C How is she doing? (POEX00047) (48) Ni je pi wa-webtawat? ni jE Opi CH.wi- webEta\AI -wad what when CH.FUT- start -35. C When are they going to start? (.TT : 3 : 5 1 : 7) (49) "Ni je zhi wej-bkedeyen ?" ni jE zhiw CH.wEjE- bEkEde -yEn what there reason. why- be.hungry\AI -2.C Why are you hungry? (JS.4.2.022) 67 Yes-no questions, on the other hand, are similar in form to the corresponding statement, taking a main clause independent verb, with a second position question particle: (50) Gdebsemen ne ewi-piekeygo? gE- debEsa/e -mEn nE e - wi- pieke -yEgo 2- have.enough\AI -15.1 Q FCT- FUT- make.pie\AI -15. C Do we have enough (berries) to make a pie? (JT. 03. 037. 008) Related languages show variability in the use of the changed conjunct with content questions. In Ottawa, for example, Valentine reports that "questions of location that do not involve a relative root do not show initial change" (2001, p. 983). In Potawatomi, a relative preverb is added, and the verb shows initial change: (51) Ni pi je ga-je-toyen? ni pi jE CH.gi- EjE- Et -o -yEn where CH.PST- where- put.s.t.\TI -OBJ -2.C Where did you put it? (JT.03. 13.009) (52) Ni pi je eje-tek? ni pi jE CH.EjE- te -g where in . a . certain . direction- be . in . a . certain . place -O.C Where is it? (JT.03. 13.007) The use of the changed conjunct in wh-questions likely reflects the fact that wh-questions trigger presuppositions, whereas yes-no questions do not. Or more precisely, 'why are you hungryV presupposes 'you are hungry', whereas the yes-no question, 'are you hungry' carries only the vacuous presupposition 'either you are hungry or you are not hungry' (Levinson, 1983). As a context for presupposition, the changed conjunct is not unexpected here (as with completed adverbial clauses discussed 68 in Section 3.2 with example 29), and is likely grammaticalized in wh-questions for precisely this reason. 4.4.4 Unaccompanied main clause conjunct Hockett (1948a) reports that the conjunct can also be used alone to express a wish, as in (53): (53) Byat! bya/e -d come\AI -3.C If he would only come! Rather than using this construction to express a wish, speakers today generally prefer to use either of the particles bedo or begesh as in (39) and (40) above. There are, however, other uses of a main clause conjunct without a particle. As with other main clause conjuncts that co-occur with a particle, these utterances imply that the speaker is taking an attitudinal stance with respect to the proposition. For example, someone might say (54) if the addressee wasn't gone as long as was expected (the addressee might respond with something like, 'well, I didn't get a chance to see the doctor'): (54) O, e-gi-gish-odankeyen? o e - gi- gizh- odanEke -yEn oh FCT- PST- finish- go. to. town -2.C Oh, you finished everything in town? (POEX00251) In (55), the speaker expresses his excitement over a fast car ride by using the conjunct, which injects a certian vividness (this sentence was translated by the speaker as 'we were going to beat hell!'): 69 (55) O, e-yapich-bozyak ! o e - yapich- boz ~y=ig oh FCT- to . such . an . extent- take . a . ride\AI -15. C How fast we were going! (POEX00263) In (56), a teasing folk saying, the speaker suggests that the unusual act of the addressee's cutting wood caused a weather event: (56) E-gi-mneseyen, wi ye i e - gi- mEnEse -yEn wi ye iw FCT-PST-cut .wood\AI-2C EMPH FRED that . INAN wech-gmeyamgek . CH.wEjE- gEmEya -mEgEg CH.the. reason.why-rain\II-AUG.O.C You cut wood; that 's why its raining! (POEX00259) When asked, the speaker would also accept an independent verb in the main clause, but explained that it didn't have the same force as a conjunct, that somehow the implication that the act caused the rain was not as strong. (57) Ggi-mnese, wi ye i gE- gi- mEnEse wi ye iw PST- cut .wood\AI . I EMPH PRED that . INAN wech-gmeyamgek . CH.wEjE- gEmEya -mEgEg CH.the . reason. why- rain\II -AUG.O.C You cut wood; that 's why its raining! (POEX00260) 4.4.5 Summary The fact that main clause conjuncts are found (sometimes grammaticalized) with particles that express propositional attitude suggests that the conjunct is being used in a subordinate context, only that the subordinator is a particle rather than the typical propositional attitude predicate. However, this argument cannot be maintained exactly as such when presented with examples such as those in the previous section which do not have a subordinating verb or particle. These examples suggest that the important aspect 70 for the use of the conjunct is the expression of speaker subjectivity, whether or not this is overtly expressed by a particle. When this is available contextually, it acts as a functional subordinator and the attitude is indirectly registered by the use of the main-clause conjunct. 4.5 The Conversational Construction (CC) This chapter has outlined the uses of independents and conjuncts in everyday discourse. While independents are always used in a main clause, conjuncts are found in both subordinate and main clauses. If we take the subordinate clause use of the conjunct to be its basic use, then we can explain its main clause use as signalling functional subordination to an implied propositional attitude. The preverb e-, which becomes important in the narrative behavior of the conjunct, has its basic use in everyday discourse as a marker of factivity. It is found only in non-hypothetical subordinate clauses: in complement clauses, it expresses speaker confidence — probability versus possibility; in adverbial clauses, it is not used in hypothetical clauses including clauses expressing futurity. I will call this basic distribution of the independent, conjunct and preverb e- the Conversational Construction (CC), to distinguish it from the pattern of independents, conjuncts, and the preverb e- which will be found in narrative discourse (discussed in Chapter 6). 71 Bloomfield, L. 1927. Notes on the Fox Language. International Journal of American Linguistics, 4.181-219. — . 1946. Algonquian. Linguistic Structures of Native America, ed. by H. Hoijer, 85-129. New York. — . 1958. Eastern Ojibwa: Grammatical Sketch, Texts and Word List. Ann Arbor: University of Michigan Press. Costa, D. 1996. Reconstructing Initial Change in Algonquian. Anthropological Linguistics, 38.39-72. Goddard, Ives. 1990. Aspects of the Topic Structure of Fox Narratives: Proximate Shifts and the Use of Overt and Inflectional NPs. International Journal of American Linguistics, 56.317-40. Hockett, C. 1948a. Potawatomi I: Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi III: The Verb Complex. International Journal of American Linguistics, XIV. 139-49. James, D. 1983. Simple versus Conjunct Verbs in Moose Cree: Some Whys and Wherefores. Actes du Quatorzieme Congres des Algonquinistes, ed. by W. Cowan, 345-61. Ottawa: Carleton University. Levinson, Stephen C. 1983. Pragmatics: Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press. Rhodes, Richard. 1985. Eastern Ojibwa-Chippewa-Ottawa Dictionary. Berlin: Walter de Gruyter & Co. Thompson, Sandra and Longacre, Robert. 1985. Adverbial clauses.vol. II Complex constructions: Language typology and syntactic description. Cambridge: Cambridge University Press. Valentine, J. Randolph. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press. 72 5 Verbal Paradigms and Mental Space Construction in Everyday Discourse 5.1 Introduction This chapter contains a Mental Spaces theory analysis of the syntactic structures discussed in Chapter 4, including the use of independents, conjuncts, and the e- preverb in everyday discourse. By determining their basic function within mental space networks for everyday discourse, we can then compare their use in a more complex network that contains an embedded narrative (this is discussed in Chapter 7). The structure of the chapter is as follows: Section 5.2 presents the case that independent verbs structure Space R, whereas conjunct verbs always structure a space embedded within Space R. Section 5.3 shows that main clause conjuncts are not problematic for this analysis, as they too occupy an embedded space, even in the absence of an overt space opener. Section 5.4 shows the use of the e- preverb is a marker of factivity of an embedded space. 5.2 The basic use of independents and conjuncts 5.2.1 Independents Independent verbs in the present tense structure Space R. Consider the following sentence: 72 (1) Nde-nna ' ikanen ni wasechgenen. nEdE- nEna'ikan -En niw wasechEgEn -En 1- fix.s.t.\TI -05.1 those. INAN window -PL I'm fixing the windows. (In response to the question 'What are you doing? ') (.TTB3. 050.0 15) This would be represented by a single space R, which houses the BASE, V-POINT, FOCUS CONTENT and FOCUS CONTEXT: (2) 'Fm fixing the windows.'^ Space R: a* \ w J "REALITY" SPACE a: first person pronoun w: windows nna 'ikek 'fix' a(w) BASE V-POINT FOCUS CONTENT FOCUS CONTEXT Sometimes an independent verb will occur in a space subordinate to Space R, such as when the verb is marked with a past or future tense. In the following example, the space o^Qnerwesnago 'the day before yesterday' opens a past space: (3} O, wesnago ge ni gbe-gizhek o wEsnago ge nin gEbe- gizhEg oh day . before . yesterday also I . EMPH through . all . of- day ngi-monshkwe . nE- gi- monshkwe 1- PST- weed\AI.I 'Oh, the day before yesterday, I weeded [my garden] all day. ' (JTB3.036.009) ^ Potawatomi verbs in mental space diagrams will be cited in the conjunct form. 73 This past Space P is subordinate to Space R. The use of the past tense morpheme gi- on the independent verb ngi-monshkwe 'I weeded' signals that focus has shifted to Space P:^ (4) 'I weeded all day' Space R: BASE V-POINT FOCUS CONTEXT Space M: "REALITY" SPACE a: first person pronoun T FOCUS CONTENT PAST FACT prior to R TIME SPACE wesnago 'yesterday' monshkwet 'weed' a The use of the independent in the example above does not open the new space, rather it is the time adverbial that is the space opener. Neither do the independent inflections /'er se indicate that the new space is in focus; this is accomplished by the tense marking.^ ■ Connectors between counterparts are not drawn in the following diagrams m order to simplify the representations. The same letters used in different spaces represent counterparts. ^ Fauconnier notes that "[tjenses and moods do not by themselves explicitly set up spaces, but they give important grammatical cues concerning the spaces relevant for the sentence being processed" (1985, p. 33). 74 5.2.2 Conjuncts Conjunct verbs, unlike independents, are indicative of a subordinate, embedded space in the mental space network. I will illustrate this by discussing two kinds of subordinate clauses, complements and adverbials. Complement Clauses. The space-building properties of certain complement taking predicates have been examined in the Mental Spaces theory literature. Fauconnier discusses the space opening properties of the predicates believe, hope, claim, (1985) want, wish, not believe, and doubt (1997, p. 95). Also, Cutrer (1994) has a detailed discussion of utterance predicates such as say. There is good reason to think that sentential complement-taking predicates in general are space openers. The propositions expressed in complement clauses usually describe an alternate world. This might be an unrealized world, as in the case of desideratives (want, wish, desire, hope), or pretense predicates {imagine, pretend, fool into thinking). The proposition might also represent the mental world of particular experiencer, as with utterance predicates {say, tell, promise), propositional attitude predicates {believe, think, assume, doubt), or 'factives' {regret, be sorry, discover, know, forget). Some predicates combine the two; in the case of predicates of fearing {fear, worry, be afraid that) there is the description of a possible state of affairs, and the speaker's mental attitude toward that state."* Main clause verbs marked with the independent can be space openers if they are complement taking predicates. Example (5) illustrates the use of the complement taking predicate, 'see'. The main clause verb ngi-wabma 'I saw him/her' takes the independent. '' Categories and examples of sentential complement-taking predicates are from Noonan (1985). 75 and the subordinate clause verb e-gishnenat 'that he/she buys it (animate)' is in the conjunct: (5) Ngi-wabma Mani e-gishnenat niw dabyanen. nE-gi- wabEiti -a mani e- gishEnEn -ad niw Odabyan -En 1- PST- see.s.o\TA -DIR.I Mary FCT-buy . s . o\TA-3/0 . C that . OBV car -OBV I saw Mary buy the car. (POEX00068) The main clause predicate wabmat 'see' opens a content space (Space N), which is occupied by the complement predicate gishnenat 'buy' . The use of the conjunct for the complement predicate indicates that the subordinate space is then in FOCUS: (6) SENTENTIAL COMPLEMENT-TAKING PREDICATE: 'I saw Mary buy the car' Space R: BASE V-POINT FOCUS CONTEXT Space M "REALITY" SPACE a: first person 'F PAST FACT prior to R Space N PAST SPACE wabmat 'see' a[ ] CONTENT SPACE m: Mary d: dabyan gishnenat 'buy' m(d) FOCUS CONTENT In this example, the paradigmatic inflection of the conjunct is not itself the space builde; the new space is opened by the sentential complement-taking predicate 'see'. This can be compared with the use independent in (4), which also did not open the subordinate space. Paradigmatic inflections are not necessarily themselves space 76 builders. Rather, they provide additional cues to the structure of the network at any given point in the discourse. Adverbial Clauses. Like complement clauses, adverbial clauses typically open new spaces. Standard examples include time spaces {in 1929), and domain spaces for works of art or literature {in that painting, in War and Peace), and hypothetical spaces (// it rains tomorrow). Example (7) contains an example of a conditional sentence. The protasis gishpen bonimgek 'if it's snowing' contains a conjunct verb: (7j Gishpen bonimgek, nwi-we-zhoshk ' o . gishpEn boni -mEgEg nE- wi- wE- zhoshEk'o if snowMl -AUG.O.C 1- FUT- go. and- go . sledding\AI . I If it's snowing, I'll go sledding. (POEX00021) The "^diXiicXQ gishpen 'if is a space builder, which opens a hypothetical Space M in a new domain subordinate to Space R, and a future prediction space N whose information is evaluated from the V-POINT of the hypothetical Space M. Space M houses the protasis gishpen bonimgek 'if it is snowing'. The use of the conjunct verb form bonimgek signals that the hypothetical Space M is in FOCUS. 77 (8) ADVERBIAL CLAUSE: 'If it is snowing. 'REALITY" SPACE Space R: BASE V-POINT FOCUS CONTEXT V-POINT PRESENT FACT not prior to R FOCUS CONTENT HYPOTHETICAL SPACE bonimgek 'snow' FUTURE PREDICTION SPACE Space N 5.3 Conjunct verbs in main clauses The fact that conjunct verbs can occur in main clauses would appear to be an exception to the generalization that conjunct verbs are indicative of a relationship of subordination. However, as argued in Section 4.3, the subordinate form of main clause conjuncts reflects a functional subordination to either an implied propositional attitude, or one that is expressed by a particle. In Mental Spaces theory terms, this propositional attitude (expressed or implied) opens a new space. , which in turn takes a complement space. The complement space houses the propositional material being evaluated from the higher propositional attitude space. Because information is being evaluated from this space, it is the locus for 78 V-POINT. The use of the conjunct signals that the subordinate complement space is in FOCUS. The function of a particle in opening a prepositional attitude space is illustrated in (9) below. The prepositional attitude space is a wish space, opened by the particle bedo 'wish that' . It takes a complement space, which houses the content of the wish. This is expressed by the conjunct verb gmeyamgek '(if) it rains': (9) Bedo (wi) na gmeyamgek. bedo wi na gEmEya -mEgEg wish. that EMPH EMPH rain\II -AUG.O.C I wish it would rain! (POEX00262) "REALITY" SPACE a: first person Space R: BASE PROPOSITIONAL ATTITUDE SPACE bedo 'wish that' a[ ] Space M V-POINT COMPLEMENT SPACE paradigm: conjunct mood: /active - b: third person inanimate gmeyamgek 'rain' b Space N: FOCUS In (10) below, the prepositional attitude is not overtly expressed. The only indication that the proposition is being evaluated in some way is the use of the conjunct verb form byat '(if) he comes', which signals the addressee to look for an evaluation. 79 When an evaluation (or the indication of an evaluation) is available in the context of the utterance, but is not overtly expressed, it can nevertheless serve to open a propositional attitude space. I call this type of contextual cue an implicit space opener (indicated in the space diagram by the use of a dashed-line text box). Note this also means that particles, as well as grammatical predicates can serve as space openers. (10) Byat! bya/e -d come\AI -3.C If he would only come! "REALITY" SPACE a: first person Space R: BASE PROPOSITIONAL ATTITUDE SPACE, IMPLICIT a: first person WISH a[] Space M V-POINT COMPLEMENT SPACE paradigm: conjunct mood: /active - b: third person byat 'come' b Space N: FOCTIS The mental space structures in (9) and (10) provide a model for sentences like (1 1) which contain particles that are not necessarily space openers. The particle anake is polysemous; it is commonly used non-evaluatively as the disjunctive 'or', but can be used evaluatively, as illustrated in (1 1) — ^the alternatives are construed as what the addressee is doing, alongside what the speaker thinks he should do. The evaluation available in the 80 context ('I think you should wait until it rains') serves as an implicit space opener. This space is then occupied by the particle anake, which is semantically compatible, and then becomes associated with the evaluative reading. Such structures likely serve as a means of grammaticalization for the use of the conjunct with such particles. (11) Gwi-gwdemo j ge ne? Anake (zhe) bama gmeyamgek. gE- wi- gOdEmojEge nE anake zhE bama gEmEya -mEgEg 2- FUT- fish\AI.3.I Q or EMPH wait rain\II -AUG.O.C Are you going fishing? Maybe wait until it rains. (POEX00258) "REALITY" SPACE a: first person Space R: BASE IMPLICIT SPACE anake 'maybe' a[ ] Space M: V-POINT Space N COMPLEMENT SPACE second person bama 'wait (until)' b[ ] COMPLEMENT SPACE c: third person inanimate gmeyamgek 'rain' c 5.4 The preverb e- Within subordinate clauses, Potawatomi has a mood distinction. Unlike languages that mark irrealis (for example, the use of the subjunctive in French), Potawatomi marks realis-type clauses by the use of a verbal prefix e-. I have glossed this prefix as 'factive' (FCT) as it has many properties of a marker of factivity, although to be 81 more accurate, it reflects the relative strength of an assertion. For this reason, the same predicate may take a conjunct marked with a factive, or not as the sentences in (12) and (13) show. Note that rather than being a property introduced by the space opening verb 'think', the feature "Factive" is a property of the complement clause (compare the case of the subjunctive in French which is required by certain predicates). Thus, if the proposition expressed by the subordinate clause is considered to be factual or probable, the verb will take the prefix: (12) [Ndenendan Mani e-wi-gishnenat l-thmk.thus .of .s .t\TI-OBJ-3/0I Mary FCT-FUT-buy . s . o\TA-DIR-3C niw wdabyanen . ] cc that.OBV car-OBV / think that Mary will buy the car. [POEX00040] However, if proposition expressed in the complement clause is considered to be only probable, the verb takes the prefixes da-je- instead, glossed here as 'modal' (MOD).^ (13) [Ndenendan Mani da- j e-gishnenat l-thmk.thus .of .s .t\TI-OBJ-3/0I Mary MOD-MOD-buy . s . o\TA-DIR-3C niw wdabyanen .] cc that.OBV car-OBV / think that Mary might buy the car. [POEX00049] I represent this distinction in mental space diagrams by use of the feature Factive {+/-}, as shown by (14) (Factive +) and (15) (Factive -): ' These are also used for deontic modality, as in 'Mary ought to buy the car. ' 82 (14) 'I think that Mary will buy the car. REALITY SPACE a: first person Space R: BASE FOCUS CONTEXT THOUGHT SPACE nendek 'think' a[ ] Space M V-POINT PRESENT FACT not prior to R Space N: COMPLEMENT SPACE paradigm: conjunct mood: /active + m: Mary d: dabyan 'car' gishnenat 'buy' m(d) FOCUS CONTENT (15) 'I think that Mary might buy the car.' "REALITY" SPACE a: first person Space R: BASE FOCUS CONTEXT THOUGHT SPACE nendek 'think' a[ ] Space M: V-POINT PRESENT FACT not prior to R COMPLEMENT SPACE paradigm: conjunct mood: f active - m: Mary d: dabyan 'car' gishnenat 'buy' m(d) Space N: FOCUS CONTENT 83 5.5 Summary The following, then, are the basic functions of verbal paradigmatic morphology with respect to Mental Space networks: In the absence of a space-builder or other linguistic cues which might indicate a special context, an independent verb will structure Space R. The use of the conjunct signals a shift to an embedded space. This space may be opened by a sentential complement-taking predicate, or may be opened by virtue of an adverbial clause. The preverb e- indicates the factivity of the embedded space in relationship to its parent space. In Chapter 7, these uses in everyday discourse will be compared with the structures they help build in narrative. 84 Bibliography Cutrer, Michelle. 1994. Time and Tense in Narratives and Everyday Language, University of California: Ph.D. dissertation. Fauconnier, G. 1985. Mental Spaces. New York: Cambridge University Press. — . 1997. Mappings in Thought and Language. Cambridge: Cambridge University Press. Noonan, Michal. 1985. Complementation. Complex Constructions, ed. by Timothy Shopen, 42-140. Avon: The Bath Press. 85 6 Independents and Conjuncts in Narrative Discourse 6.1 Introduction In this chapter, I examine the use of independents and conjuncts in narrative discourse, where they have a very different distribution from their use in everyday discourse. Narrative discourse is marked by the high frequency of verbs inflected in the conjunct, which occur in main as well as in subordinate clauses. These conjuncts are usually preceded by the preverb e- (I will refer to this preverb-verb combination as an "e-conjuncf ). This is illustrated by the excerpt given in (1), (verbs are underlined, the use of brackets and the notation "CC" and "NC" are explained below): ^ (1) 1 [I me se ngodek neshnabek e-wdodanwat i je Once there was a Village (some weye e-nshonaj tagwat wgetkansewan mine people had a village) and someone mbish wed'emwat.] was destroying their gardens and their wells. 2 [Iw je nish wshkabewsen e-gi-nokanawat So they had two SCOUts watch out for e-wi-kewabmawat wegwendek o ezhcheget.] whomever might be doing that. 3 [I je bama zhe na geten e-byanet weye.] Later, sure enough, someone came along . ^ Examples contain line numbers to the left of the Potawatomi text, which reference the line numbers in my translations. Line numbers are referenced here as (example:line), as in (1:5). Verbs in the Potawatomi text are roughly indexed to the English translation with underlines. Where the sequence of verbs does not match, numeric indices are given. The code in parentheses after the last line of the English translation indicates the source text. Interlinear glosses of examples are given in Appendix B. 85 4 [ E-wabmawat kojesen e-bshkobnanet ; jak zhe They saw him pulling out beans and na e-zhechgenet.] doing all kinds of things. 5 Wabozoyen je ni. It was Rabbit. (JS.4.1) Together, the preverb and conjunct form a construction characteristic of narrative, particularly the genre of mythological narrative called yadsokanen .^ Hockett proposed that this construction functions to "[set] the style of the text, which is a story, not supposed necessarily to be true, at least, not intended as a recounting of anything which once happened to the narrator" (Hockett, 1948d, p. 216).'' The use of the e-conjunct is usually established in the first sentence of a yadsokan, which, along with the optional but common formula (i me se) ngodek 'once', functions to announce the narrative performance, as shown in (2) - (4) below: " The other main narrative genre, yajmownen, includes autobiographical and historical texts. These narrative types, which are not included in the current corpus, need to be considered independently. ^ Other Central Algonquian languages, such as Ottawa (Nishnaabemwin) and Fox, show a similar use of the conjunct in main clauses for narrative discourse. In Ottawa, the parallel construction is the plain (unchanged) conjunct. According to Valentine, "[t]he reason. . . is simply that sentences in running narrative sometimes act as if they were subordinated to the whole narrative, or form tight units with adjacent sentences" (Valentine, 2001, p. 951). In Fox, a similar construction takes a conjunct verb preceded by the cognate preverb e-(h)- (glossed as 'aorist'). According to Bloomfield, "[t]his is the commonest form of the conjunct; in hearsay narrative it replaces the independent mode of ordinary speech." (1927, p. 204) Although Potawatomi is more closely related to Ottawa, speakers of Potawatomi and Fox shared a more recent period of close contact which resulted in many lexical borrowings from Fox into Potawatomi. In this case, it is Fox construction and not that of Ottawa which appears to be the closest to Potawatomi, and may in fact be the source for the Potawatomi construction in its modem form. 86 (2) 1 [Ngodek wabgonoshkwe e-gche-mwet jik- zibe. ]nc (3) 1 [ode yadsokan espen e-bmebtot . ], (4) Once a rat was cryiiiR by the edge of a river. (H0PT2) This story is about the Raccoon running along . (JS.4.4) 1 [I me se ngodek neshnabek e-wdodanwat . J^c Once there was a Village. (Literally: 'Some people had a village .') (JS.4.5) A main clause verb in the e-conjunct, as well as any subordinate clauses forms a grammatical pattern which I will call the Narrative Construction (abbreviated NC)."* The Narrative Construction contrasts with the Conversational Construction in the form of the main clause verb, as shown in (5). A reasonable analysis might limit the domain of the construction to the main clause, defining the distribution of independents and conjuncts per se. This, in fact, has been the traditional analysis. However, an argument for including subordinate clauses in the construction comes from the behavior of main clause verbs of speech, where the paradigmatic form of the main clause verb imposes an interpretation on the content of the direct speech in the subordinate clause (this is described in §3.1). Also, the construction is limited to a single main clause verb and any subordinate clauses: verbs in juxtaposed or conjoined main clauses can belong to different construction types, as shown by 24:29 and 24:31. As will be argued below, the CC and NC constructions are associated with different discourse functions, and this domain for the construction (main plus subordinate clauses) is proposed (at least for Potawatomi) as the minimum unit with which these discourse functions can be associated. 87 (5) A COMPARISON OF THE NC AND CC Construction Type Main Clause Verb Subordinate Clause Verb(s) Narrative Construction (NC) e -conjunct (e-) conjunct Conversational Construction (CC) independent (e-) conjunct This statement of the contrast between the NC and CC requires quaUfication. First, it is unclear whether conjuncts in subordinate clauses inflect, taking the preverb e- or not, just as they would in conversation. Part of the difficulty in determining this with certainty is the rarity of contexts in narrative clauses that would require a conjunct without e-. There are no examples of hypothetical clauses outside of direct speech in the corpus, and only three instances of 'before' clauses, two of which show contradictory treatments, given in (6) and (7). In (6), the narrator uses an e-conjunct in the adverbial 'before' clause, which goes against conversational usage (see Chapter 4, examples 24 - 25); and in (7) a different narrator uses a conjunct without the preverb, in accord with conversational usage: (6) 50 [ E-hvamsiie -nyetrgongsk e-byawat giw neyap i je o nene e-nat niw osen, "Nnedwendan debendemak . " ] nc Before the four days ended , the couple came back , and the man said to his father, "I want our belongings.' (JS.4.2) (7) 46 [Iw je i ga-nakwneget e - wi - debma t pi bwamshe gwabtonet . ] ^c The one that planned it would grab him before he reached the shore . (MD 102694) 88 There are many sociolinguistic factors which could potentially account for this difference: the speakers are grew up in different communities, belong to different generations, and show idiolectal variation in narrative style. There is also the potential factor of using ofbwamshe 'before' as a preverb in (6), and as a particle in (7).^ At this point, there is simply too little data to suggest an analysis. The second qualification concerns narrative sentences with main clause conjuncts that appear without the preverb e-. There are a few such sentences in the corpus; examples are given in (8) and (9), which are both from the same text: (8) 28 [A^ , baJjwichget j igbyek . ] nc Ah, he waited there by the shore. (MD 102694) (9) 35 [A, gkanabmat o wabozo.li Ah, the Rabbit looked across at him . (MD 102694) Since both verbs are imperfective, and the sentences appear in different parts of the story, it is likely that this is some other construction type, rather than a production or There is the additional complexity of (6) and (7) belonging to different narrative discourse types. (6) is a narrative sentence, and (7) is an explanatory aside, which, as will be discussed below, have different grammatical requirements. "^ There is a preverb a- that appears infrequently and in similar contexts as e-. However, the intonation and pauses in the recording of this text indicate that the a in (7) and (8) are clearly interjections rather than preverbs. (The interjections a and o are frequently found at the beginning of sentences in this text, and as is often the case with interjections, their semantic contribution is difficult to pin down). 89 transcription error. As with the case of adverbial clause usage, more data will have to be analyzed before this can be worked out. Abstracting away from these complications, we will say for now that the primary difference between the NC and the CC is the form of the main clause verb. This contrast becomes important in narrative, since, although the NC is the predominant construction found myadsokanen, there are usually several instances of the CC in any given text, sometimes occurring in sequences of sentences. According to Hockett, independent verbs in narrative (that is, instances of the CC) indicate "explanatory material directed to the listener, not integrally part of the story, or else direct quotation" (Hockett, 1948d, p. 216). Indeed, throughout the texts, direct speech always occurs in the conversational pattern. This is illustrated in (10) by the speech of two characters, Rabbit and Lion. (Sentences in the NC are indicated by surrounding the clause in brackets followed by a subscript "NC" label, and sentences in the CC are indicated by the use of brackets followed by a subscript "CC". If there is no finite verb in the main clause, as in the case of verbless sentences (see 1:5), or when the main clause verb is a participle (see 14:7), the construction type is formally — although not necessarily functionally — indeterminate, in which case, no surrounding brackets are used.) 90 (10) 44 [Iw je e-bme-byat niw beshkmwen e-nat] Nc [ "Nseze ! Gyetnam nzeges. ]cc 45 [ Nwebi ' we . ] cc 46 [Weye zhode nshiwnagze anwe ge gin gneshiwnagwes nesh je win nwech.]cc 47 [ Ibe gge- zhyamen ; geten nshiwnagze ■ " ] cc 48 [Beshkmwe e-kedot , ] f,c [ " Gzhyamen , gge-we-wabmamen . " ] cc When he [Rabbit] came across the Hon he said to him, "Brother, I'm very scared . I'm running away from someone. Someone here is pretty scary ; and you're scary , but he's even worse. Let's go over there; he sure is scary ." Lion said, " Lets go and take a look at him." (JS.4.1) Hockett provides three examples of 'parenthetical explanation' which come from the first of two glossed texts in his sketch. These are given in (1 1) and (12) below (my transliteration, Hockett' s translations): (11) [Neshnabe je o weni'get espen gi -y awe . ] cc 'When the Indian went trapping, the raccoon went along . ' (HOPTl) (12) 11 [Gi je yaygenwik je giw; ]cc [jo je mamda e^ 'They were just the same Size, these wi-wepodwat ; e-bwa-gkenmat ni wde- two, you see; SO it was impossible for espenmen . ] ^c him [the man] to hit him [the other coon] ; he couldn't tell which one was his own. ' 12 [Pene je ni wde-espenmen nam-yegwan gi- wjeshnon. ] cc 'His own coon was always underneath. ' (HOPTl) Hockett' s analysis of the use of independents in direct speech need not be disputed, since it is uniformly the case. However, the analysis of remaining instances of independent verbs as occurring in 'explanatory material' raises several questions. One question lies in 91 defining what is meant or encompassed by 'explanatory material.' Is it the case that the CC marks background information? And if this is the case, does the NC by contrast mark foreground information, or the 'main thread' of the narrative? Hockett's analysis also raises questions of descriptive adequacy. Many instances of the CC in narrative defy categorization as explanatory material, or even inclusion in the wider category of background material. Can these instances themselves be categorized, and if so, what relationship do these uses have, if any, to uses already described? In the discussion below, I argue that the main distinction between the CC and NC is, in fact, their role in grounding (Section 6.2) and that the remaining uses of the CC can be explained as instances of narrative-internal perspective (Section 6.3). 6.2 Grounding Linguistic analyses of narrative discourse usually recognize two broad types of clause: one type which provides the main events of a narrative, and another which provides supportive information such as explanations, evaluations and descriptive commentary. The terminology for these two types varies, however, I will refer to the main narrative information as 'foreground' and the supportive information as 'background'.^ In the following sections, I show that a main function of the CC is to encode background information, and in contrast, the use of the NC in narrative encodes foreground information. ^ The use of these terms is after Hopper (1979a; 1979b) and Hopper and Thompson (1980) who compared this discourse phenomenon to a gestalt figure/ground relationship and tied it into a larger discussion of language and cognition. Labov (1972) uses the terms 'narrative clause' and 'non-narrative clause'. Grimes uses 'event' and 'non-event' 92 The discussion in this section is based on Grimes's analysis of narrative (1975) which recognizes the need to partition narrative information into these two categories. 6.2.1 Use of the CC for background According to Grimes, background information includes settings, explanations, and evaluations. Each of these types is discussed in turn below. Settings. Settings include information about the time, place, and location of a narrative, or give information about the circumstances in which a narrative takes place (Grimes, 1975). The excerpt in (12) below contains an example of a setting. After the opening sentence, the storyteller switches to the CC. The reason for the shift is to provide information that sets up events in the story: (13) In the story of Raccoon and Wolf, Raccoon knows where a stash of pork rind is, and while out on his forays, has also found a beehive. In the first episode of the story. Raccoon tricks Wolf into thinking the beehive is the sack of meat. The following information prepares the listener for the setup of the trick: 1 [Ode yadsokan espen e-bmebtot . ]hc This story IS about the Raccoon running along . [E-ye-bmebtot o espen wgi-wabman amon While Raccoon was running along , 2 e-gojnenet . ] cc he saw bees (a hive) hanging (from a tree). [Ga-zhewebzet je gi-gmegmode gokosh He would go about stealing pork 3 wzheyen ngo j i . ] CC rind somewhere. (JS.4.4) Some texts, like that of the example just given, dispense with the setting in a matter of one or two sentences. Other texts have several sentences at the beginning which serve as an setting. In the following excerpt, the setting begins at line 2, and runs through line 6 (and 93 arguably through line 7, although the discourse pattern of line 7 is not discernable). The narrative proper begins at line 8, which switches to the NC. The NC continues then as the predominant pattern: (14) [I me se ngodek neshnabek e-wdodanwat . ] i [Gi-dbedbowek ; gego zhena gi- yajdanawat . ] ^c [I je ibe iribesek nawesh [gagita] odan gi -y awen ibe . ] cc I je ye i ga-wje-dbedbowewat . 5 [I je ngot nene neshzhena gi-wijewe neko . ] cc 6 [Jo zhena win gego gi-zhe-dbowesi neshzhena e-zhyat . ] cc 7 Ga-wj e-zhyat je e-wi-mnekwet . 8 [Ngodek e-dokit bama zhena j o weye ; ] nc [ j ayek gi-maj iwagben . ] cc 9 [E-gingenayek nsheke . ] ^c 10 [Ngodek jigbyek e-gi-we- j a j ibdebet gdewanen e-giwadzet i je o mtek e-gi- ggenonat . ] ^c Once there was a village. [More literally, 'some people had a village']. They were having a council ; talking about something. And there was a town in the middle of a lake. That's where they would go for their council. So there was one man who used to go along for no particular reason. He did not go for the counc il; he went for no particular reason. The reason he went was to drink . Once this man woke up and nobody was there; everyone must have left . He was left all alone . One time he went by the lake and sat by a log, feeling lonely , and the tree spoke to him. (JS.4.5) Explanations. Grimes describes explanations as "not part of the narratives themselves, but [information that] stands outside them and clarifies them," and that ".. .explanations and comments about what happens have a secondary role that may be 94 reflected in the use of distinctive grammatical patterns" (Grimes, 1975, p. 55-6). In Potawatomi, explanations are marked by the use of the CC, which sets them off from the majority of the narrative sentences in the NC. For example, the last clause in (14:8) (which occurs after the setting) explains that the man suddenly finds himself alone because his friends have abandoned him. It is common to find such sentences in the CC occurring in isolation within a narrative. This is probably because explanations generally have a local function, serving as asides that comment on or explain events in nearby sentences. Settings, in contrast, tend to be longer and generally occur at the beginning of a narrative; their location is in keeping with their more global function of providing information which helps stage the narrative as a whole. Examples like (13:8) which provide additional information about the story -world are what I call story-internal explanations. They are fairly common in the corpus, and include Hockett's examples of 'parenthetical explanation' in (1 1) and (12). Additional examples are given in (15) - (19) below, preceded by a description of the context: (15) A village chief has been trying to get Rabbit killed by sending him on all kinds of perilous missions. None of these devices work, and in the end, it is Rabbit who kills himself by following through on a boast that he can walk through a fireplace without harm. Of course, a fireplace isn't very perilous unless there is a fire in it, so the narrator takes pains to interrupt the story in order to provide the fire: ^ Grimes uses the terni 'background' for what I am calhng 'explanations'. I reserve the term 'background' to refer to the broader category that includes settings, explanations and evaluations. 95 88 [I je iw bodwagen megwa shkode gi^ So there was still a fire in the temget . ] cc fireplace. (.TS.4.1) (16) A man is out hunting with his wife and son. The woman, in gathering bark to make cord, meets a bear with whom she initiates a sexual relationship. In Algonquian lore, animal-human matings upset the natural balance which can lead to all kinds of trouble, providing plenty of fodder for stories.^ In this tale, the man's hunting is affected, and he cannot kill anything. The man ends up near starvation, but the woman and the boy are well-fed and happy. As an aside, the narrator posits the following as the reason for their different situations: 9 [Ode mko wgi-sheman . ] ec This bear was feeding them. (JS.4.6) (17) In the French Story, a destitute boy and his grandfather are able to raise their fortune as a result of being taught blacksmithing by the French Spirit. In the process of acquiring stock, they obtain a pony that turns out to be magical. The narrator explains the special function of the pony in lines 18 and 19: 18 [O negdoshas wgi-nizhokmagwan . ] cc The pony helped them. 19 [ E-bwamshe-je-yewawat negdoshayen Before they had the pony, the wgi-wbesh'eg wan seksiyen wgetganeswa. ] cc deer were rummg their gardens. (JS.4.3) (18) A boy and his grandfather discover a scheme to spy on them, cooked up by the man's son and the son's wife. The couple hide her mother in a box, provisioned with food, and leave the box of 'valuables' with the boy and grandfather to guard while they leave to go on a trip. The boy and grandfather discover the old lady in the box, which they have been using as a dinner table. Line 48 provides the prop which the boy uses to suffocate the old lady (line 49) while she is unconscious. ' This insight is from Richard Rhodes (p.c). 96 4 7 [Iw je e-gi-babgemat niw ndemozeyen. ] i 48 [Jak bkwezhgasen wa-mijet zhiw gi- tene. ] cc 49 [Iw je e-gi-bkwenshkodwat niw bkwezhgasen mine iw ziwabo abte e-gi- zigwebnek . ] uc So he knocked the old lady (in the head). All the crackers for her to eat were there . He stuffed the crackers in her mouth and poured out half of the cider. (JS.4.2) (19) The last example comes from the story of Raccoon and Wolf. After Raccoon and Wolf get to a stash of meat inside a shed, Raccoon selects a piece and drags it back out the hole where they crawled in. Wolf, however, gorges himself all night, which explains why he was unable to scramble away when the white people come into the shed to check on their meat: 39 [O mwe gi-wzam-debsenyet jo mamda e-wi-majnewit e-pich-dbomayek . ] qq The wolf was too full ; he couldn't move away while they talked over (what to do about) him . (.TS.4.4) Sometimes, a narrator will refer to a cultural practice in order to explain story events. 10 which I call story-external explanations. Two examples are given in (20) and (21). (20) A listener in hearing the opening of the Rabbit Story (see example 1), might object that the villagers, angry at Rabbit for destroying their gardens, would just kill Rabbit outright. If they could, of course, we wouldn't have much of a story. To counter this potential objection, the narrator interjects a reference to background cultural knowledge: community law prevented the villagers from executing the Rabbit, which is why they tried to set up his 'accidental' death: It is worth noting that the narrator's audience, Hockett, was an outsider to this community, and presumably was not familiar with these cultural practices. 97 8 [Iw je o wabozo zhiw gi-dbendagze odanek Since the Rabbit belonged to the jo je mamda i e-wi-zhe-nsawat mamwech village, they couldn't kill him as bshe gego gj iyek bama a-je-nsawat . ] cc they please; they would have to get something more on him m order to kill him . (JS.4.1) (21) In the story of the woman who has relations with a bear, the son, who wants to tell his father what is happening, is prevented from doing so because his father is out hunting during the day, and the boy sleeps with his mother at night. The narrator provides cultural information to explain why the husband and wife slept separately. This information also reinforces why the husband's hunting was affected by his wife's behavior: success in hunting is attributable largely to following certain codes of behavior. A man and wife sleeping together during the hunt is enough to affect hunting success, let alone the extraordinary situation of one's wife sleeping with a bear. 10 I je iw pi neshnabek e-giwsewat jo And when people went hunting, [wgi-widpemasiwan wdekweyomwan babkan ^jjgy ^j^jj^'^ gjggp ^j^j^ tj^gjj- ^ives; zhena gi-nbek . ] cc t^ey slept separately. (JS.4.6) Evaluations. Evaluations are clauses that express the speaker's reaction to events in the narrative, or to the narrative as a whole. Evaluative clauses can occur throughout narrative (Labov and Waletzky, 1967), and tend to be mobile, that is, they can be extracted and placed at other points in the narrative without significantly disrupting the narrative continuity (Grimes, 1975). In the Potawatomi narratives I have examined, evaluations tend to occur at the beginnings and ends of narratives, often in thematically paired sequences of sentences where the sentences in the conclusion recapitualate those of the introduction. This seems to be a common phenomenon with stories whose telling serves an explanatory or 98 moralistic function: as Grimes notes, "a story with a moral is. . .likely to be an exhortation within which there is an embedded narrative" (Grimes, 1975, p. 64). The excerpt in (21) is from a modern text that 'explains' why rabbits today have short tails. Lines 1-5 contain the initial evaluative material. The narrator returns to this theme in line 56 after the conclusion of the main narrative 11 (22) [O, neko ngi-babzedwak neshnabek e-yayajmowat eyayengajmowat . ] cc I used to listen to the people telling stories ; something they laughed about. 2 3 4 5 [Iw je] ni wabozoyen ngodek e-gi-yajmawat . Once thev told abou t Rabbit. [O, bnewi neko o wabozo gi-gnewanwe . ] cc [ Gi - gnewanwedek kedwik . ] ^c Oh, at one time Rabbit had a long tail. He must have had a long tail , they say . Iw je i wech-shkwanwat ngom ga-zhewebzet . That's why he has a short tail today, because of what happened to him . Continued... 56 Iw je iw yedek wech-ngom-shkwanwat o wabozo, [ gi -kedwik neko gi gekyajek neko e-gi-wnanodogwa e-yangajmowat . ] cc That's must be why Rabbit has a short tail today, the elders used to say , when I heard them telling funny stories . (MD 102694) The French story, given in (23) - (25) and discussed below is a similar example, having extensive thematically related evaluative sections. Labov (1972, p. 371) notes that narrators sometimes stop in the middle of narration to address the listener and tell what the point of the story is. He calls this 'external evaluation', since it is a break from the storytelling frame. The example in (22) would fall under Labov's category of 'embedded evaluations', a more sophisticated device which does not break the continuity of the story. 99 6.2.2 Use of the NC for foreground If a primary function of the CC in narrative is to indicate background, then one must next address whether the NC is used for foreground. In order to see if this is the case, we will now examine the French Story, a short narrative given in its entirety in (23) - (25) below. This narrative is an example of a story told as an explanation for a real world phenomena. As discussed above, a story which functions as an explanation commonly has evaluation sections which bracket an embedded narrative. In this case, the embedded narrative tells the story of how the French Spirit helps out a destitute boy and his grandfather. The evaluation sections explain that some Potawatomi cultural practices are ultimately attributable to the French (via the French Spirit). The story begins with an evaluative section (lines 1-13) which is told almost entirely in the CC. Most of the verbs are imperfective, and the clauses are not temporally ordered: (23) 1 [Ngom wdopi wemtegozhi yewak naganit.]cc 2 [Iw je ngom wdopi nnodamen weye e-wepodek biwabek wizhgya e-nayek wi zhe ibe Kansas memek . ] qq 3 O je ye o gche-mnedo eng[e]t wemtegozhi. 4 O ye o gangezot wemtegozhi ekdonegek . 5 [ I j e ngom bme-yewak zhena nekmek . ] cc [Jo win gdemagzesi ginan wi eneshnabewigo gdekdomen . ] cc [Wemtegozhi mane ton wzaw-zhonya mine mkede-biwabek . ] qq [Mine ngom e-gkendemgo bgoch-negdoshayek mine seksik jak zhena e-yemgek . ] uc [Ode je nene win wdebenman . ] cc Up to today, the French are the leaders somewhere. Nowadays we hear someone blacksmithing , especiaUy there in Kansas, they say . That's the great spirit of the French. That's the lost French, so they say. Now he is moving around in different places. He is not poor ; we who are Indians say that. The French have lots of gold and black iron. And today we know wild horses and deer and so forth are there . This man owns them. 100 10 [Ode je wemtegozhi gzhe-mnedon Now God helped the French to be wgi-nizhokmagon e-wi-mishgwezet ode je powerful , but our brother the Spanish anet gikansenan Spanish e-nayek , ode was Victorious ., they saVi. wdekwenzhgewan 2 . ] cc 11 [I je ngom wdopi ode wemtegozhi nwech zhe Up to today, the French are very ninweze zhode kik . ] cc weak m the world. 12 [Ngodek je ode wemtegozhi wgi -ni zhokmowen At one time, the French helped out neshnaten . ] cc the Indians. 13 [I je i pi ode wemtegozhi wgi-minan At that time the French gave him a ngemwen i je ye i ngom gode neshnaJDek song, and that's the one these Indians e-yewat i je ode ngom nim'ediwen gode here use m their dancing to this day . neshnatek e-gche-yowat . ] cc (JS.4.4) The switch to the NC in Une 8 is at first surprising, since it seems to be a fi'ee clause just Uke the surrounding sentences. ^^ However, it is different in that it takes place in 'real' time, as opposed to 'story' time. It is structurally similar to line 2, which also is framed as the present with ngom 'now, today'. However, the reference to wild horses is based in reality (there were, for example wild horses on certain Potawatomi reservations within people's memory) compared to the blacksmith of line 2, which seems to represent a mythical or spiritual being. The function of line 8 seems to be an aside, making it an aside within the larger evaluative section which is in itself a kind of aside. Since the CC is expected in evaluations, perhaps the preferred way to distinguish such 'double asides' is to switch into theNC. The next section contains the narrative proper. This begins at line 14, where the storyteller switches to the NC. The NC is used throughout this section to form the matrix of sequential events in the story. The sentences that occur in the conversational pattern (indented here from the other text) are background information. Like the clauses in the ^' The first verb is a conjunct, since it has the e- preverb rather than stem-internal change expected of the participle. The mam and subordinate clauses are therefore in the NC. 101 opening evaluation, these sentences tend to be non-sequential (lines 19 and 30) and frequently contain verbs with imperfective aspect (lines 18, 25, and 29): (24) 14 [I je o wemtegozhi e-gi-nat niw gigabeyen] nc [ "Nasena zhechgen ezh-widmonan . " ] cc 15 [I je o wemtegozhi e-wishteyaywat i e-gkeno ' mewat2 ni gigateyen. ] nc 16 [I je o gigaJDe wikapi e-gi-ne-wishteyaywat e-gi-gkeno ' mowat niw wmeshomsen . ] nc 1 7 [Wikapi e-gi-negdoshayensawat m ine zhena gego . ] ^c 18 [O negdoshas wgi-nizhokmagwan. ] cc 19 20 25 [ E -bwamshe - j e -ye wawa t negdoshayen wgi-wbesh ' egwan seksiyen wgetganeswa . ] cc [Gigate e-ggenonat ni negdoshayen, ] ^c [ "Ni je wa-zhechgeyan ? " ] cc 21 [I je o negdosha e-nat , ] nc ["Wigbish mtegok wdenen ge-dkobdon nkwegnak gekwedso ' egme-kezhyep ge-giwta ' omgon iw ggetganwa . ] qq 22 [I je gi seksik e-wi-zegzewat . ] cc 23 [Nesh je gego zhe gwi-zhe-ngok ■ " ] cc 2A [o seksi e-kedot ] nc [ "Wegni je o Wakayaiide bye-zizdeyatek? " ] cc [Egme-kezhyep zhena o je wemtegozhi ni zhokmo wen i je mine wa-mijwat ^ wiyas o gi-wje-wdetnanawa2 . ] cc 26 [Ga-gish-jagnenet wdenwemagnen wmeshomsen ga-gish-ihbonet e-gi -ma j i t . ] ^c The French told one boy, "Be careful to do things the way I tell you to." So that French (Spirit) was teaching s the boy how to blacksmith i . Finally, the boy started to blacksmith , and he taught his grandfather. Finally, they had a pony and so forth. The pony helped them. Before they had the pony, the deer were ruining their gardens. The boy asked the pony, "What should I do ?" And the pony said , "Get some bark from the basswood tree, tie it around my neck, jump on , and ride me around your garden every morning. The deer will be scared . Of course, they will say something to you." The deer said "What does that round- tooth have sticking out between his legs?"'^ Every morning the French Spirit helped them, and that's how they obtained ^ their meat to eat i . After his relatives and grandfather died, he left. ^ Native speakers are unsure exactly how this sentence should be translated. It may be a sexual joke, or it may refer to the monstrous appearance of a man ridmg horseback. Round-tooth may be an epithet for a human being (as used by the deer!). 102 21 [I je geyafce wemtegozhi nizhokmowan geyabe j e ngom gnizhokmagnan . ] cc 28 I je ngom pi neshnate wemtegozhi mskwe we j -gwgezhkek o wemtegozhi e-gi-zhwenmat . 29 [Ode gigate e-gi-majit e-gi-byat odanek neshnaben eyenet;]nc [ga-gkendek je ni wemtegozhiyen wgi-gkeno 'mowan neshnaben wa-zhi ' enet . ] qq 30 [Ga-gish-gkeno 'mowat wiznaben wa-zhetonet, gego wgi-nan : ] cc ["Gego nsedi ' kegon , " ] cc [ wgi-nan. ]cc Still the French helped him, and is helping us to this day. Up to today Indians have French blood inside them, because the French (Spirit) blessed them. This boy left and came to where there was an Indian village; what he learned from the French he taught the people who were there. After he taught his fellow people what to do, he told them something: "Don't kill one another," he said. 31 [I je o wemtegozhi e-gi-nat niw gigabeyen ga-widmowak e-wi-bwa-mje-dodadwat , ] f,c [mine i je ngom wdopi neshnabek enwek-dbenbwek . ] cc And the French told the boy what to tell them, that they should not abuse each other , and so up to this day, the Indians are surely civilized. (JS.4.3) In the conclusion of the story, the narrator returns to the evaluative theme of the introduction, reiterating the reason for the story's telling. Once again we have the evaluative information coded in the CC. (25) 32 [I je ngom wdopi dewe ' gen-nim' ediwen debwetanawa neshnabek i ye i ga-gowat ni wemtegozhiyen . ] cc 33 [Iw je ngom wdopi jak neshnabek wdebwe t anawa ode madmowen iw je we j -mno-widokwdadwat . ] cc 34 [Nchiwenmok e-wabdawat ngom wdopi . ] cc Up to this day the Indians believe in the drum dance; that's the one the French told him about. And up to today, all the Indians believe this way and that's why they are good friends. They are happy to see each other up to today. 35 Iw je ekwak ode wemtegozhi yajmowen. So that's how long this French story is. 103 6.2.3 The grounding function of the CC and NC Based on the data presented above, it seems clear that a primary function of the CC and NC in narrative is to distinguish foreground and background information/"* It is no surprise that Potawatomi should grammaticalize a grounding contrast. It has been proposed that the foreground/background distinction is a functional universal in narrative discourse (Hopper, 1979b). Nor is it surprising that such a contrast should be achieved by means of morphological marking on the verb: languages show considerable variation in the grammatical devices which they employ to encode grounding; these range from the use of specialized discourse particles to the verbal properties of aspect, voice, and even word order (Hopper, 1979a). In some languages, such as English, grounding isn't associated with any single grammatical feature, but rather is associated with a set of properties (Hopper and Thompson, 1980).*^ ^'' It might be suggested that the CC and NC are not being used for grounding at all, but are rather the result of a process like clause chaining. In languages that use clause chaining in narrative, a series of non-finite clauses is terminated by a finite clause. The function of clause chaining appears to be to delimit sentences by topic, since each finite clause corresponds roughly to the end of a paragraph (Longacre, 1985, p. 265). Under a clause chaining analysis we would therefore expect a more even distribution of independent verbs to reflect thematic divisions in the text. However, as we have seen, independent verbs do not have an even distribution; in fact, at first glance they appear to have a scattershot distribution except m the introductions and conclusions of texts, where they cluster (due to their use for settings and evaluations). ^^ Foreground clauses are associated with high transitivity, with verbs that tend to be perfective, sequential, kinetic events and realis; background clauses are associated with low transitivity, with verbs that tend to be imperfective, non-sequential, stative and irrealis (Hopper and Thompson, 1 980). 104 6.3 Use of the CC for internal viewpoint Once we redefine the primary use of the CC in narrative as encoding background information, we take care of many instances of the CC, notably settings and evaluations, which cannot be perspicuously defined as 'explanatory material'. However, several puzzling instances remain which defy even this wider categorization/^ In this section, I argue that these problematic examples show the use of the CC for narrative-internal viewpoint, where the narrator represents information as coming from a particular character's point of view. This imposition of a unique perspective different from the narrator's can have the effect of making the narrative more lively: the audience 'sees' through the character's eyes.^'' A primary function of internal viewpoint is therefore for vividness. However, because internal viewpoint limits vision to the character, it can also be used to restrict the validity of information to that character. Along with the function of vividness then, another function of internal viewpoint is to emphasize the epistemic distance between the narrator's thoughts and beliefs, and those of a character. This analysis finds support in the use of the CC for direct speech, which has also been shown to be a kind of internal viewpoint. In a study of news texts, Sanders and Redeker ^'^ A likely explanation for why Hockett missed these problematic cases was that the texts he collected contain relatively few instances of the CC outside of direct speech. Modem texts that were first audiotaped and then transcribed indicate a much more frequent use of the CC, and therefore many more instances outside of direct speech which require an explanation. " I use the term 'internal' perspective in contrast with 'external' perspective, where narrators report the actions of characters. This is the classical distinction between mimesis and diegisis (Plato, 1 968). This topic has received considerable attention in the field of narratology, where it is also refered to as 'focalization' — see Genette (1980) for a discussion. 105 (1996) found that internal perspective is an important function of reported speech: with indirect speech, the narrator shares responsibility for the content with the subject, however in direct speech, the responsibility is presented as remaining entirely with the subject. Therefore, treating the examples below as cases of internal perspective subsumes them under the broader umbrella of perspective phenomena that includes direct speech, allowing what would otherwise be problematic instances of the CC to be easily assimilated into the present analysis. Section 6.3. 1 contains examples of the CC used for vividness. Section 6.3.2 shows the use of the CC for epistemic distance, in a particular context I call 'quote frames'. Section 6.3.3 shows other cases of epistemic distancing. In Section 6.3.4, 1 argue that the use of the CC for epistemic distancing has been extended to a new context, what I call 'semantic opposition' . 6.3.1 Vividness Internal viewpoint can be used so that the narrative seems to come from a particular character's point of view. This has the effect of making the narrative more lively; the narrator 'shows' what happened instead of reporting it. In the story of How Rabbit Got a Short Tail, the narrative begins with the Rabbit stopped on the shore of a river, wishing to cross it in order to get to the clover on the opposite side. In line 12, the Crocodile character is introduced. Line 13 is in the NC. In line 14, the narrator switches to the CC, apparently taking the rabbit's perspective, since what is 'sticking out' is most apparent to an observer above the water: 106 (26) 12 I je ge wi zhi o gagtanago i yedek. So must be Crocodile was there. 13 [Beshoch zhe na zhi jigbyek [ge] He was floating m the water near the e-gegwijek . ] nc shore . i-^ [ z agw j anegwi j en zhi . ] cc His nose was sticking out there. (MD 102694) In the next example, we again see through the Rabbit's eyes, since the crocodile is only in 'last place' if he is located at the opposite shore from Rabbit: (27) A6 [[win] ibe shkweyak gi-nshkweshen i ga- The Crocodile that planned it lay at nakwneget gagtanago .] cc the end , there m last place. (MD 102694) The use of the CC for vividness seems to be less common among the 1940's texts, although the following are two possible examples (the story of Raccoon Running Along), where the viewpoint in line 28 is the Raccoon's, and the Wolfs in line 29: (28) 28 [Espen o mtegok gi-gdegosi e-wawabmat niw The Raccoon was high (in a tree) and mwen wete zhe e-gi-bdek ■ egaznet . ] cc saw the Wolf get badly stung . 29 [I je o mwe jo gi-nshkadzesi ; neshnege That Wolf didn't get mad ; he still megwa gi-dnendan i wa-zhyawat e-wi-gmodwat thought the meat would be there, and goko shen . ] cc wanted to go there and steal that pork. (JS.4.4) What makes it difficult to decide on a vividness analysis for examples like (28) is that they could also be explained as instances of explanations. While it is difficult to tease these two analyses apart, the fact that most potential 'vividness' examples show this dual 107 interpretation could be an added motivation for the development of the CC as a perspective device. 6.3.2 Quote frames and epistemic distancing As we have seen, the conversational pattern is always used in narrative to represent the speech or thoughts of a character. In (26), an excerpt from the story oi Raccoon and Wolf, the discourse of the two characters (lines 6-10) takes place in the CC.^^ In the larger sentence which embeds each quote, the verb of speech is in the conjunct, indicating the use of the NC which is consistent throughout the larger passage (as shown by the inclusion of lines 5 and 11): (29) 5 [Gete zhena e-gi-nkweshkwat mwen.]} Sure enough, he (the Raccoon) met Wolf. 6 [ "Nshi , gde-ton ne gego wa-mijyan ? " ] cc "Brother, do you have anything to [e^nat espenen.]Nc eat?" he said to the Raccoon. 7 ["Jo zhe kwech bkeji nde-ton wa-mijyan "Not much, I JUSt have a Uttle toeat nawkwek , " ] cc [e-nat espen . ] nc foj- ^y gwn dinner," said the Raccoon. 8 [Mwe e-natewat , ] sc [ "Wegni je etoyen ? " ] cc Wolf asked him, "What do you have?" ^^ The use of the CC for direct speech also extends to multiply -embedded quotes, where characters report the speech of other characters. In the following example, both the narrator's and the character's quotations are in the CC: 20 [ Epitajmewat ngot mine e-kedot , ]xc ["Shebzhi ngi-nek, [ ' Nin nda-nsa, ' ] cc kedo. " ]cc While they were talking , another man said , "Lion said to me 'I can kill him' [he said]." (JS.4.1) 108 9 [Espen e-nat , ] ^c [ "Mteno zhe na bkeji gokosh-wzhey ndesa, " ] cc [e-nat . ] ^c Raccoon said , "I have just a little meat-rind," he said. 10 [Mwe e-nat , ] nc [ "Mo jma shemshen o wzhey • " ] cc 11 [I je o espen msach e-gi-minat . ] nc Wolf said , "Please feed me that rind." So the Raccoon finally gave it to him. (JS.4.2) While the use of the NC to frame quotations appears to be the norm; it is not universally the case, as shown by the example from the Hard Life story in (30). In lines 64- 66, it is the CC and not the NC which frames the quotations: (30) 62 [ Ga-gish-ngo ' wawat gigate neyap e-wawidmewat niw keweziyen, ] nc [ "Nmesho , ngodwak gwkengo ' gazo o ndemoze . " ] cc 63 [ Nwi -mon ' wa . " ] cc 64 [Kewezi]cc "Jo, gego" [wdenan. ]cc 65 [ " Gda-bon-gdemagzemen iw ngodwak, "]cc [wdenan. ] cc 66 [Kewezi]cc "Gego" [ wdenan . ] cc In the story of How Rabbit Got a Short Tail, used for a quote frame: (31) 27 ["A, iw zhe yedek e-wi-dkemozh ' ewat gode , " ] cc [ zhede ' e o wabozo.]cc After they buried her , the boy went back and excitedly told the old man, "Grandfather, one hundred dollars is buried with that old lady. I'm going to dig her up ." But the old man said , "No, don't." "We could quit living poorly with that hundred," he said to him. "Don't," said the old man. (JS.4.2) we find a similar example of the CC "Ah, must be they [the Crocodiles] will tak e me across," thinks the Rabbit 19 (MD 102694) In Potawatomi narrative, reported speech, including the inner speech of thought, is typically represented as direct speech. Potawatomi has indirect speech, however, outside of narrative. 109 However, two sentences earlier in the same text, we have the following minimally distinct example, with the Rabbit's thoughts framed in the NC: (32) 25 ["Gego zhe ode gagtanago nwi-nakwnek ,""[00 [e-zhde ' at o wal30zo.]f,c "This Crocodile has something planned for me ," thought the Rabbit. (MD 102694) A few lines later in the same story, we have another example of Rabbit's inner speech framed intheNC: (33) 36 "O, wzam ne zhe gete ode? 37 [Gagtanago nwe j itmagodek ? " ] cc [s- zhde ' at . ] ^^ "Oh, can this really be? Will Crocodile really help me?" he thought . (MD 102694) What both (32) and (33) appear to have in common is Rabbit's suspicion of Crocodile's intentions. These stand in contrast with (3 1), where Rabbit thinks Crocodile and his cronies will help him out. In the latter cases, Rabbit's suspicion is in accord with the beliefs of at least the narrator and probably the audience as well, who likely come to the story with expectations about the Crocodile's dubious character. In (3 1), however, we have the contrast of Rabbit's naivete; an epistemic state which the narrator represents as distant from her own. The analysis that the CC is used by narrators for epistemic distancing finds support in the otherwise problematic instances of the Crocodile's speech in the How Rabbit Got a Short Tail story (lines 15 and 19), where the quotes are framed in the CC: 110 (34) 15 ["A! Nshi ! Ni je ezhwebzeyen ? " ] cc [ wdenan ni wabozoyen. ] cc (35) 19 ["O, jo wi zhe na gego abje yawsenon i , " ] cc [kedo o gagtanago . ] cc "Ah, little brother! What' s the matter ?" he said to the Rabbit. (MD 102694) "Oh, there's nothing much to that," said the Crocodile. (MD 102694) These can also be analyzed as epistemic distancing, since the narrator and audience are unlikely to have empathy for the Crocodile character. Returning again to the example in (30) (repeated below), the reported speech in lines 64-66 framed in the CC may also represent internal viewpoint. Here however, there seems to be a shift: the contrast is not between the epistemic state of the narrator versus the character, but rather between the characters themselves, who hold conflicting points of view. (36) 62 [Ga-gish-ngo ' wawat gigabe neyap e-wawidmewat niw keweziyen, ] no [ "Nmesho , ngodwak gwkengo ' gazo o ndemoze."]cc 63 [Nwi-mon ' wa . " ] cc 64 [Kewezilcc "Jo, gego" [ wdenan . ] cc 65 [ " Gda-bon-gdemagzemen iw ngodwak ,"] cc [wdenan. ] cc 66 [Kewezi]cc "Gego" [ wdenan . ] cc After they buried her , the boy went back and excitedly told the old man, "Grandfather, one hundred dollars is buried vyith that old lady. I'm going to dig her up ." But the old man said , "No, don't." "We could quit living poorly with that hundred," he said to him. "Don't," said the old man. (JS.4.2) 111 6.3.3 Other cases of epistemic distancing We now turn to examples other than quote frames which show the use of the CC for epistemic distancing. In the story of How Rabbit Got a Short Tail, as Rabbit is running across the bridge created by the crocodiles' backs, we are told (using the CC) that coming from his perspective ('if someone were to see it'), there appears to be a hole in the water (the narrator later described it as the entrance to a burrow). The audience, of course, knows that it isn't a hole at all, but Crocodile's gaping jaws, waiting to grab Rabbit: (37) 48 [O, [nme pa zho] megwa e-gche-bmebtot bama Qh, as he was dashing across , he zhe gate... [o] bikwa zhe na wangoyane soon [saw something] that looked JUSt wiye gego e-wabdek . ] cc ^xk^ ^ hole, [more literally: it was just like a hole when somebody saw It]. (MD 102694) Any character can serve as the locus of viewpoint in a story, including the narrator in the past. In (38), which comes from the end of the How Rabbit Got a Short Tail story, the narrator tells a mini-narrative about when she saw rabbits as a child and believed their tails had really been bitten off She begins in the NC (line 59). In line 60, she restricts the viewpoint to her thoughts as a little girl, switching to the CC to show the epistemic contrast with her current adult knowledge. She evaluates this belief from an adult perspective in line 61: 112 (38) 59 [Iw je o wabozo neko e-wi-wabmek megwa e-penojewyan , iw zhe neko i e - gwdenmewek iw wzewangos . ] nc 60 [O, gete zhe na ode gi -gi shk j egadenek iw wzewangos , neko ngi-zhde ' a . ] ^c 61 Nmet se na yedek wi na ! And when I used to see the rabbit, when I still was a child , I used to feel his little tail. Oh, for sure that little tail was bitten off , I used to think . I don't know about that! (MD 102694) 6.3.4 Semantic opposition The last set of examples from the corpus that are subject to an internal perspective interpretation are shown in (39) and (40): (39) 28 [I je ode nene e-gi-nme-ninwezet e-wzam- bkedet , ] mc [i j® ode kwe mine o gigabe pene zhena winwa gi-gimoch-wisnik . ] cc (40) 116 [I je gi wewiwdeyek e-gi-yewat jayek debendemwat e-gi-mbomgek e-gi- gdemagzewat . ] f,c 117 [Mine wzhonyamwa e-gi- jagsanek . ] ^c 118 [O je kewezi mine gigabe gi-mno- bmadzik. ] cc So this man got to be weak from hunger , but the woman and the boy were secretly eating . (JS.4.6) And the couple settled; all that they owned [their stock and fowl] died, and they were poor . Also their money ran out . But the old man and the boy lived happily . (JS.4.2) These examples have similar adversative semantics, comparing the opposite situations of the protagonist and antagonist. Although the participants whose situation is framed in the CC changes (in (39) it is the antagonist's whereas in (40) it is the protagonists'), in both cases the second situation mentioned is the one framed in the CC. 113 Also, in both cases, the character(s) mentioned in the first part of the comparison are the ones that have been the subjects of the immediately preceding discourse. It is possible that this adversative-like use of the CC could have developed out of the use of internal perspective for epistemic distancing, with the intermediate step of examples with quote frames that contrast the mental opposition of two characters within the story. From this point, it is but a short leap in use to contrast the opposite situations of those characters. These three uses are contrasted in examples (41) - (43) below: (41) EPISTEMIC DISTANCE BETWEEN NARRATOR AND CHARACTER (repeated from (31)) 27 ["A, iw zhe yedek e-wi-dJcemozh ' ewat gode , " ] CC [ zhede ' e o wabozo . ] cc "Ah, must be they [the Crocodiles] wiU tak e me across," thinks the Rabbit 20 (MD 102694) (42) EPISTEMIC DISTANCE BETWEEN CHARACTERS (repeated from (30)) 62 [Ga-gish-ngo ' wawat gigabe neyap e-wawidmewat niw keweziyen, ] nc [ "Nmesho , ngodwak gwkengo ' gazo o ndemoze . " ] cc 63 [Nwi -mon ' wa . " ] cc 64 [Kewezi]cc "Jo, gego" [wdenan. ]cc 65 [ "Gda-bon-gdemagzemen iw ngodwak , " ] cc [ wdenan . Ice 66 [Kewezilcc "Gego" [ wdenan . ] cc After they buried her, the boy went back and excitedly told the old man, "Grandfather, one hundred dollars is buried with that old lady. I'm going to dig her up." But the old man said , "No, don't." "We could quit living poorly with that hundred," he said to him. "Don't," said the old man. (JS.4.2) "° In Potawatomi narrative, reported speech, including the inner speech of thought, is typically represented as direct speech. Potawatomi has indirect speech, however, outside of narrative. 114 (43) OPPOSING SITUATIONS OF CHARACTERS / ADVERSATIVE (repeated from (39) 28 [I j e ode nene e-g±-nme-ninwezet e-wzam- bkedet , ] nc [i js ode kwe mine o gigabe pene zhena winwa gi-gimoch-wisnik . ] qq So this man got to be weak from hunger , but the woman and the boy were secretly eating . (JS.4.6) 6.4 Summary In the preceding sections, I have identified the uses of the NC and CC in narrative as shown in (41). As compared with the single discourse use of the NC for foreground clauses, the CC presents a rather large array of functions. The analysis presented above suggests grouping these into two main discourse contexts: background and internal viewpoint. (44) USES OF THE NC AND CC IN NARRATIVE Narrative Construction (NC): Conversational Construction (CC): 1 . Foreground clauses 1. Background: a. Settings b. Explanations i. Story-internal ii. Story -external c. Evaluations 2. Internal viewpoint: a. Direct Speech b. Outside of direct speech i. Vividness ii. Epistemic distance between narrator and character iii. Epistemic distance between characters iv. Semantic opposition / adversative As a marker of foreground clauses, it is not surprising that the NC is the most common construction encountered in narrative. In contrast, it is surprising that the less frequent CC should occur in such a wide variety of narrative contexts. A possible series of historical developments that could explain these various uses of the CC is outlined below. 115 6.5 Possible historical sequence of CC uses in narrative It is likely that the first step in the development of the various uses of the CC in narrative was its use to represent direct speech. Here the CC is clearly iconic for basic conversation; we construe characters' dialog in a story as a kind of conversation, based on our understanding of how conversations work in reality. At this point, by virtue of its use to represent direct speech, the CC could become associated with internal viewpoint. Presumably, the reported conversation of characters in a story is normally used for vividness,^^ so it is likely that this was an early use of the CC outside of direct speech. However, internal viewpoint naturally extends to the representation of epistemic distance, allowing the CC to extend to these contexts as well. The primary use of the CC for epistemic distancing appears to be a contrast between the narrator and character's point of view. However, we have also seen cases where this is extended to represent opposing points of view between characters in a narrative, as in (30). Once the construction comes to represent a contrast contained within the bounds of the narrative, it is a short step to its use as an adversative, as in examples (39) and (40). Thus we have the following hypothetical series of developments: "' There is no choice of direct or indirect speech m Potawatomi narrative, at least, one never finds indirect speech. However, a narrator can choose to report what characters say or simply describe their actions. 116 (45) Basic mode Direct speech (internal viewpoint) Internal viewpoint: ► • For vividness For epistemic distance Epistemic Distance: _^ - Between narrator and character -► • Between Characters >1 O fV O ^ Adversative ^5<^ 1=^ go With this analysis, once we estabUsh direct speech as primary among the uses of the CC in narrative, the development of the other uses follow in a straightforward fashion. Although the beginning and endpoint of the series (direct speech and adversative uses) are quite different from each other, the stages in between represent rather small semantic changes. 117 (Hockett, 1948a) (Hockett, 1948b) (Hockett, 1948c) Bibliography Bloomfield, L. 1927. Notes on the Fox Language. International Journal of American Linguistics, 4.181-219. Genette, Gerard. 1980. Narrative Discourse, an Essay in Method. Ithaca: Cornell University Press. Grimes, Joseph E. 1975. The Thread of Discourse.vol. 207: Janua Linguarum Series Minor. The Hague: Mouton. Hockett, C. 1948a. Potawatomi I: Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi II: Derivation, Personal Prefixes, and Nouns. International Journal of American Linguistics, XIV. 63-73. — . 1948c. Potawatomi III: The Verb Complex. International Journal of American Linguistics, XIV. 139-49. — . 1948d. Potawatomi IV: Particles and Sample Texts. International Journal of American Linguistics, XIV.213-25. Hopper, Paul. 1979a. Aspect and Foregrounding in Discourse. Discourse and Syntax, ed. by Talmy Givon, 213-41. New York: Academic Press. — . 1979b. Some observations on the typology of focus and aspect in narrative language. Studies in Language, 3.37-64. Hopper, Paul J. and Sandra A. Thompson. 1980. Transitivity in Grammar and Discourse. Language, 56.251-99. Labov, W. and J. Waletzky. 1967. Narrative analysis. Essays on the Verbal and Visual Arts, ed. by J. Helm, 12-44. Seattle: Unversity of Washington Press. Labov, William. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Longacre, Robert. 1985. Sentences as combinations of clauses. Language typology and syntactic description, ed. by Timothy Shopen, 235-86. Cambridge: Cambridge University Press. Plato. 1968. The Republic. New York: Basic Books. Sanders, Jose and Gisela Redeker. 1996. Speech and Thought in Narrative Discourse. Spaces, Worlds, and Grammar, ed. by E. Sweetser. Chicago: The University of Chicago Press. Valentine, J. Randolph. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press. 119 7 Mental Space Construction in Narrative 7.1 Introduction In this chapter, I present a Mental Spaces analysis of the use of the CC and NC in narrative discourse. By using this model, we are able to capture the difference between the use of these constructions, as well as similarities across the various uses of the CC in narrative. The discussion is based on the work of Cutrer (1994), who analyzes the use of tense in written narrative using mental spaces. I show that this analysis is helpful for Potawatomi, but requires some modification to accommodate oral narrative. I also argue for an elaborated representation of ground in Mental Spaces theory. 7.2 The domain of narrative A narrative event is represented by the creation of a narrative Space N which is set up relative to Space R. The embedding of the narrative space within Space R reflects that narration takes place within the larger context of speaker "reality".^ Any of several grammatical as well as non-verbal cues (attention getting devices, special seating arrangments, etc.) can serve to open the narrative space. Potawatomi has ^ Here, I am referring to a traditional narrative, rather than narratives that are told in a few sentences in everyday discourse. Although the latter type of narrative is not explictly addressed here, those I have examined take the form of everyday discourse, and use the CC. I assume that traditional narrative is a marked form of discourse, both m function and form. If, or to what extent, this is also the case of casual narrative in everyday discourse is the subject of further study. 118 an explicit narrative space building phrase: / me se ngodek. .. (or a minor variation of this phrase) which functions like the English 'once upon a time' . The switch to the NC, which often takes place in the first sentence, can also signal the beginning of a narrative. Throughout the course of a narrative, multiple spaces will be created subordinate to Space N. These spaces might be past spaces, future spaces, hypothetical spaces — the same kinds of spaces that are opened in everyday discourse, only they are happening within the context of the narrative. These spaces, along with Space N, constitute a narrative domain, separate from the spaces set up in the reality domain, which include Space R and its other daughters.^ " I take 'domain' to mean a partition of spaces, used to group spaces that constitute potentially alternate construals of reality. Other examples of domains may be found in Cutrer (1 994), and include hypothetical domains set up by the protasis of conditional sentences, as well as the representation of alternate viewpoints in direct speech and narrative. 119 (1) REPRESENTATION OF THE NARRATIVE DOMAIN Space R "REALITY" DOMAIN NARRATIVE N4, etc. The narrative domain brings with it a V-POINT (represented in (1) with the symbol "@"). The V-POINT in the "ReaUty" Domain is that of the speaker; in the Narrative Domain, the V-POINT is that of a fictional narrator. The concept of fictional narrator is based on Cutrer's analysis of written narrative as containing multiple V-POINTs, including a domain for implied author (supplied by the frame of novel writing), and another for a fictive narrator/narratee (evidenced by the 120 ''parcours du recit"^ where the narrator and narratee are observers within the narrative). This model is too elaborate for oral narrative, which does not motivate an intervening 'implied author'. However, when speakers make use of a narrative-internal perspective (such as presenting the narrative from the viewpoint of a particular character), I will argue that they access the viewpoint of a fictive narrator in the Narrative Domain. 7.3 Grounding As discussed in Section 6.2, Potawatomi grammatically differentiates foreground and background sentences by the use of the NC for foreground and CC for background. In this section, I argue that the use of these grammatical constructions reflects a difference in the mental space configurations for foreground and background. 7.3.1 Foreground I will begin my analysis of foreground information by examining the opening sentence of a narrative, given in (2) below. Both main clause verbs evidence the use of the NC (main clause e-conjuncts are underlined): (2)6:1^ 1 [I me se ngodek neshnabek e-wdodanwat i Once there was a Village (some je weye e-nshonaj tagwat wgetkansewan people had a village ) and someone mine mbi sh wed ' emwat . ] ^^^ destroymg their gardens and their wells. (JS.4.1) ^ The term is from Fauconnier (1984). ^ The examples given here are repeated from Chapter 6. These numbers refer to the example number in Chapter 6. The glosses for these examples are provided in Appendix B. 121 The phrase /me se ngodek, along with the NC serves to open the narrative Space N. FOCUS shifts to the embedded Space N, which is structured by the events and characters of the story. The basic function of the NC is therefore to signal that the Narrative Domain itself (rather than a particular space within the domain) is in FOCUS. BASE and V-POINT remain in Space R. This configuration (shown in (3)) represents an external, or objective, narrative viewpoint.^ (3) REPRESENTATION OF FOREGROUND INFORMATION 'REALITY" DOMAIN Space R: BASE V-POINT NARRA TIVE DOMAIN Space N FOCUS CONTENT This analysis of narrative foreground differs from Cutrer. In her analysis, the activity of "narration" takes place from the V-POINT of fictive narrator inside the Narrative Domain. Cutrer argues, based on Fauconnier (1984), that this latter domain is always available as a potential BASE; "it can be highly elaborated in fiction [as in the parcours du recit] ... or used in its more abstract form for everyday story-telling." By external viewpoint, I mean diegesis, i.e. the act of 'telling' (as opposed to internal viewpoint, or mimesis, i.e. the act of 'showing'). 122 Narration, then, for her, involves the relocation of BASE and V-POINT to a space inside the narrative domain. This type of vantage point seems more natural in written fiction. Since the written channel adds an additional layer of separation between the audience and the storyteller, the parconrs seems to be a means of heightening the reader's involvement by virtually placing the narrator and reader at the 'scene' of narration. I would argue that while the BASE of the Active narrator is always available, it is not the location from which oral narration canonically takes place. Rather, it seems more likely that this takes place from a BASE within the "reality" domain. The BASE and V-POINT of Active narrator will, however, be central to the representation of internal viewpoint, discussed below (see Section 7.4). 7.3.2 Background When narrators provide background information, they step out of their role as narrator to address the listener in the here and now; the activity shifts from narration to description, or explanation. In this case, my analysis also differs from Cutrer's. Because narration for her takes place from within the domain of the fictive narrator/narratee, she is able to analyze background information as a BASE shift, or return to Space R.^ This analysis will not work here, since I argue that BASE remains in the "reality" domain for both narrative foreground and background. It seems that what is at issue is not the BASE, but in fact ^ For explanatory information, she uses the term 'external evaluation' after Labov (1972) and Fleischman (1990). 123 FOCUS. Consider the following sentence containing background information (the main clause independent verb is underlined): (4) 6:20 8 [iw je o wabozo zhiw gi-dbendagze Since the Rabbit belonged to the odanek jo je mamda i e-wi-zhe-nsawat village, they couldn't loll him as ' '^'^ something more on him m order to kill him . (JS.4.1) This sentence, coded with the CC as background information, is in one sense about what is happening in the story; we learn that the Rabbit belongs to a village whose citizens have been plotting his demise. On the other hand, the sentence is also about what the narrator thinks the listener knows; in this case, about customs regarding village membership, namely that a village member cannot be indiscriminately put to death. The speaker may have fashioned this explanation anticipating an objection from his audience that the villagers would have simply killed the Rabbit outright.^ As with narrative foreground, BASE and V-POINT remain in Space R (see (5)). The primary difference between the two types of discourse is in the addition of a focused discourse participant. FOCUS CONTENT is associated with the narrative domain (attached to Space N for the sake of simplicity) because its spaces continue to be structured by the new information. However, at this point, the narrator in a sense steps outside the narrative to attend to the needs of the hearer, providing information the hearer ^ This is a likely motive given the narrative context; the primary audience was a linguist from outside the community . 124 needs in order to understand one of the premises of the narrative. Because there is attention on a discourse participant, there is a focus on the "ReaUty" Domain, particularly on the mental space that represents the hearer's conceptualization. We represent this by associating FOCUS CONTEXT With Space H, in the "Reality" domain. (Note this case is analogous to the case of a wh-question (see Chapter 3), although the mental space structure to which it applies is more complex.) (5) REPRESENTATION OF BACKGROUND INFORMATION Space R: BASE V-POINT Space H: FOCUS CONTEXT REALITY" DOMAIN NARRA TIVE DOMAIN Space N: FOCUS CONTENT 7.4 Internal viewpoint Besides the use of external viewpoint, where the narrator reports events taking place in the story, narrators often use an internal viewpoint; representing information as coming from a vantage point within the narrative itself In Potawatomi, internal 125 viewpoint is marked by the use of the CC. The uses of the CC in narrative are described in Chapter 6, but are briefly summarized here. One of the most common forms of internal perspective is the representation of the speech of characters in a narrative. Here the distinction must be drawn between indirect speech, where the narrator reports what a character says, and direct speech, where the narrator takes on the persona of the character and acts out what the character says. In Potawatomi narratives, the speech of characters is always portrayed directly.^ Sometimes narrators use an internal vantage point in order to make the narrative seem more vivid; as if the narrator and narratee were witnessing the events of the narrative take place. ^ This vantage point is arguably that of a fictive narrator (as in parcours du recit), or may be that of a character. In any case, the viewpoints of fictive narrator and character are often closely associated. Because an internal viewpoint can restrict the outlook on the narrative world to a character's point of view, narrators may also use it to emphasize the epistemic distance between a character's point of view and their own. In the rest of this section, I will present mental space configurations for several types of discourse that can be categorized as having internal perspective. These include direct speech, vividness and epistemic distance. As will be shown below, the difference between these types of internal perspective can be easily captured using the Mental ^ Indirect speech is found, however, in everyday conversation. ' This can also be used to add humor, especially when the character is not human and therefore an unexpected perspective. 126 Spaces framework. In addition, Mental Spaces theory will allow us to motivate the use of the CC across these contexts. 7.4.1 Direct speech Reported speech has recently been addressed in the mental spaces literature as part of a larger discussion of perspective phenomena (Cutrer, 1994; Mushin, 1998; Sanders and Redeker, 1996). In Cutrer' s model, which has gained general acceptance, a reported speech event opens a speech space S, which houses the speech verb itself (if explicit)^*^, and a subordinate content space, which I will call Space C (for the character). The content space and it daughters are partitioned into a speech domain, which represents the "reality" of the speaking character. The content space carries with it a potential V-POINT; that of the speaking character (represented as "@"). So in (6), if the speaking character is Rabbit, the character domain represents his thoughts, construals and viewpoint. ^° Cutrer argues that this space exists even without an explicit space-opener. Her example is interior monologue in fiction, where the inner speech of a characater is reported as direct speech, and no speech or thought verbs are used. The absence of the speech or thought verb is merely "one less cue to the BASE shift" (1994, p. 406). 127 (6) REPRESENTATION OF REPORTED SPEECH "REALITY" DOMAIN Space R NARRATIVE DOMAIN CHARACTER DOMAIN (Rabbit) Space C Consider the Potawatomi sentence given in (7). In Potawatomi narrative, the speech and thoughts of characters are typically presented as direct speech, followed by a verb of speech or thought: (7) 6:32 25 ["Gego zhe ode gagtanago nwi- nakwnek , " ] cc [e-zhde ' at o watozo.ljjc "This Crocodile has something planned for me ," thought the Rabbit. (MD 102694) I will now build the structure for this sentence as it might be temporally constructed, beginning with the quote, as shown in (8). The speech event itself supplies 128 the speech space (Space S) and the speech content space (Space C). The speech content space houses the V-POINT associated with the character domain, in this case, Rabbit's. As names of characters, wabozo 'Rabbit' Siwd gagtanago 'Crocodile' are entities which populate the narrative Space N, and counterparts are set up as needed in spaces subordinate to Space N. The information in the quote structures space C (and its daughter spaces) and sets up counterparts for the rabbit and crocodile, which are connected to Space N. Because the quote precedes the verb or speech or thought, the speech space (Space S) will be open as a placeholder before it is actually structured by the verb of speech or thought. With the quote is given, FOCUS shifts to the domain of the character. The space it attaches to is a future space (Space Ci) set up to house Rabbit's prediction, 'This crocodile has something planned for me.' This future space is set up relative to Space C. The use of deictic expressions such as the first person prefix n- indicates that BASE has now shifted to Space C. The use of the future tense indicates V-POINT has shifted to Space C as well.^^ This V-POINT represents the first person perspective of the Rabbit. ^^ According to Cutrer, "by convention, direct quotation indicates a shift in BASE and creates a strong barrier which makes speaker reality inaccessible to deictics" (1994, p. 404). 129 (8) REPRESENTATION OF A CHARACTER SPACE "Gego zhe ode gagtanago nwi-nakwnek..." 'This crocodile has something planned for me... ' REALITY SPACE 'REALITY" DOMAIN NARRATIVE DOMAIN NARRATIVE SPACE w: wabozo 'Rabbit' g: gagtanago 'Crocodile' SPEECH SPACE CHARACTER DOMAIN SPEECH CONTENT SPACE FUTURE PREDICTION SPACE nakwnat 'plan' g(w) Fomi: CC Space C 1 : FOCUS CONTENT 130 Now let us consider the remainer of the sentence outside the quote, e-zhde'at o wabozo 'the rabbit thinks (thus)'. Space S, which is already open by virtue of the speech event, is now in FOCUS as it is structured by the thought verb e-zhde'at. We are no longer in the Character Domain, but are back in the Narrative Domain. The thought verb is marked with the NC, which indicates narrative foreground; BASE and V-POINT shift back to Space R. (9) REPRESENTATION OF A SPEECH / THOUGHT SPACE ...e-zhde'at o wabozo '... thinks the rabbit ' "REALITY" DOMAIN Space R NARRA TIVE DOMAIN NARRATIVE SPACE w: wabozo 'Rabbit' g: gagtanago 'Crocodile' SPEECH/THOUGHT SPACE zhede 'at 'think' w [ ] Form: NC CHARACTER DOMAIN Space C 131 7.4.2 Vividness Example (10) below illustrates the use of the CC for vividness. In the first two sentences, the narrator describes the Crocodile's position in rough detail. However, in the third sentence ('His nose is barely sticking out'.'), we zoom in: the Crocodile is now viewed at close proximity from a vantage point above the water, as if we were looking at the scene from the rabbit's position on the shore. (10) 6:26 12 I je ge wi zhi o gagtanago i yedek. 13 [Beshoch zhe na zhi jigbyek [ge] e-gegwijek . ]nc 14 [ Z agw j anegwi j en zhi . ] cc So must be Crocodile was there. He was floating in the water near the shore. His nose was sticking out there. (MD 102694) In mental space terms, vividness is represented by a V-POINT shift from the 'Reality" Domain to the Narrative Domain: 132 (1 1) REPRESENTATION OF VIVIDNESS "REALITY" DOMAIN Space R NARRATIVE DOMAIN CHARACTER DOMAIN Space C Besides the use of the CC which sets off such sentences from surrounding foreground material, other evidence of a V-POINT shift comes from the use of deictic expressions. In (10), the choice of the verb determines the vantage point from above the water. In (12), the Crocodile is only in last place with respect to the position of the Rabbit: (12) 6:27 46 [[win] ibe shkweyak gi-nshkweshen i ga- The Crocodile that planned it lay at nakwneget gagtanago . ] cc the end , there m last place. (MD 102694) There are two possibilities for V-POINT here; a fictive narrator (the optional viewpoint which comes with the Narrative Domain), or a character within the story. 133 Much of the time, it is not possible to make a principled choice between the two. In the case of (10) and (12) above, the perspective might be the character, or the fictive narrator in the same viewing position. However, in a few cases, the observer is clearly independent of the character, as in the following example, where the jumping Rabbit is described in the third person: (13) (see Appendix B for gloss) 6 [Jigbyek ibe e-pa-zhyat . ] hc He went around there by the water. 7 ["O, begesh na ezhi gameyek gshketoyan "Oh, I wish I could make it to e -by ay an ,"\cc [e-kedot. ] nc cross over and get there ," he said . 8 [ E-dnednangedok jigbyek.] ^c He was talking to himself along the river. 9 [Gegpi zhe gwagwashkze ' o ■ ] cc Finally, he starts jumpmg up and down . (MD 102694) Some instances of vividness evidence a shift in BASE as well. In (10), the independent verb zagwjamgwijen 'have one's nose float' ^^ has no tense morpheme, which indicates that it is present tense. In the following example, however, the independent verb is marked as past tense, which means it cannot be the BASE: (14) 6:28 28 [Espen o mtegok gi-gdegosi e-wawabmat The Raccoon was high (in a tree) niw mwen wete zhe e-gi-bdek ' egaznet . ] cc and saw the Wolf get badly stung . (JS.4.4) There seems, therefore, to be a cline in the degree to which perspective shifts to a narrative internal V-POINT, which is illustrated by the three diagrams in (15). In (15a) ^' This verb includes the incorporated form for 'nose' -jane-. 134 ('Finally, he starts jumping up and down'), the viewpoint shifts to the Narrative Domain. The use of the present tense indicates a BASE shift as well. According to Cutrer, this use of tense is evidence of a cognitive association between the viewpoints of the speaker and narrator (in this case the viewpoint of the 'external' narrator in the "Reality" Domain), which she represents by a connector linking the two viewpoints (i.e. the temporal V- POINT dimension is shared by both narrator and speaker). In (15b) (The Crocodile that planned it lay at the end, there in last place'), the BASE does not shift (evidenced by the use of past tense), but now the V-POESTT is ambiguous between the internal narrator in the Narrative Domain and the character. This ambiguity represents the cognitive immersion of the discourse participants in the narrative world. I represent this by a connector between the Narrative and Character Domains, since they share the locative V-POINT dimension." In (15c) ('His nose is barely sticking out'), BASE and V- POINT shift to the Narrative Domain. Now there are two cognitive connections: the Narrative Domain shares the temporal dimension with the "Reality" Domain, but the locational dimension with the Character Domain. (See following page.) ^' Alternatively, the V-POINT could be placed in the Character domain with a connector to the Narrative domain. There does not seem to be any principled way to distinguish these two alternatives. Rather than being a shortcoming of the model, this may help explain the vividness effect as a blurring of the two viewpoints. 135 (15) TYPES OF PERSPECTIVE SHIFT A) BASE AND V-POEVT SHIFT; "REALITY" AND NARRATIVE DOMAINS LINKED 'REALITY' NARRATIVE CHARACTER BASE V-POINT FOCUS CONTENT, B) V-POEVT SHIFT; NARRATIVE AND CHARACTER DOMAINS LINKED "REALITY' NARRATIVE BASE V-POINT FOCUS CONTENT CHARACTER Example 13; Finally, he starts jumping up and Example 12: The crocodile that planned it lay at down. the end, there in last place. C) BASE AND V-POINT SHIFT; ALL DOMAINS LINKED 'REALITY' NARRATIVE BASE V-POlNT FOCUS CONTENT CHARACTER Example 10: His nose is just barely sticking out. So rather than representing vividness as a single mental spaces configuration, it seems best to characterize vividness as a set of configurations that minimally shares a viewpoint shift from the "Reality" Domain to the Narrative Domain. As will be shown 136 below, this characterization will be sufficient to motivate the use of the CC in vividness contexts. 7.4.3 Epistemic distance Besides the effect of vividness, narrators sometimes use an internal perspective to emphasize the epistemic distance between their perspective and that of a character's. In (16), when the rabbit sees what the speaker knows to be the Crocodile's gaping jaws, the narrator reports that, from the Rabbit's perspective, it would look like a hole in the water: (16) 6:37 48 [O, [nme pa zho] megwa e-gche-bmebtot Oh. as he was dashing across , he bama zhe gate... [o] bikwa zhe na soon [saw something] that looked wangoyane wiye gego e-wabdek . ] cc j^g^ ^;^^ ^ ^ole. [more HteraUy : it was just hke a hole when somebody saw it] . (MD 102694) The narrator takes pains, however, to introduce an impersonal weye 'somebody' who does the seeing. We do not see through the character's eyes, but from the same vantage point. Here is another case where the Active narrator V-POINT is closely associated with that of a character. We represent this in mental space terms similar to the way vividness is represented; by shifting V-POINT to the Narrative Domain. This is the viewpoint of the 'internal' narrator. We capture the effect of epistemic distance by assigning FOCUS CONTEXT to Space R, since we are contrasting the conceptualization of the narrator with that of the character: 137 (17) REPRESENTATION OF EPISTEMIC DISTANCE "REALITY' DOMAIN Space R BASE FOCUS CONTEXT NARRA TIVE DOMAIN CHARACTER DOMAIN The V-POINT of the fictive narrator is also utilized for epistemically distancing a speaking character. However, because the CC is needed to represent the character's speech, it cannot be used for evaluating what is said. Rather, this is marked in the narrative domain on the speech/thought verb, in what I call the quote frame. Consider the following example: 138 (18) 6:31 27 ["A, iw zhe yedek e-wi-dkemozh ' ewat "Ah, must be they [the Crocodiles] gode,"]cc [zhede'e o wabozo.]cc will take me across ," thinks the Rabbit/"^ (MD 102694) The thought verb, zhede'e is in the independent mode (underlined), which indicates the use of the CC. The narrator uses the CC here to contrast the epistemic stance of the rabbit's naivite with the speaker and hearer's knowledge of the crocodile's true intentions — that he plans to gobble up the rabbit (this example can be compared with the sentence given in (7) where the rabbit's suspicions are in accord with the narrator's and the NC is used). The use of the CC on the main verb has the resulting effect of framing the character's speech with the narrator's evaluation of it. Epistemic distance in a quote frame is represented by V-POINT and FOCUS shifting to the space for the speech/thought verb. Because this space stands in the Narrative Domain but contiguous to the Character Domain, it is a convenient place to mark evaluative information about the quote. ^^ ^'' In Potawatomi narrative, reported speech, including the inner speech of thought, is typically represented as direct speech. Potawatomi has indirect speech, however, outside of narrative. ^^Some languages (like Potawatomi) maintain the integrity of the speech content space; others apparently do not. In Cayapa, for example, a verbal suffix -n marks events that figure into role reversals for the story characters. If an important event is mentioned by a character, the verb will be marked with -«, even though the character may have no awareness of the event's significance (Longacre, 1 976). Cayapa presents a problematic case for Sanders and Redeker's (1996) analysis, which treats direct speech as having the strongest possible character perspective. They discuss four types of perspectivization phenomena: direct 139 (19) REPRESENTATION OF A QUOTE FRAME "REALITY" DOMAIN Space R: BASE FOCUS CONTEXT, NARRATIVE SPACE w: wabozo 'Rabbit' Space M NARRA TIVE DOMAIN THOUGHT SPACE w: wabozo 'Rabbit' zhede 'at 'think' w Form: CC CHARACTER DOMAIN Space C mode, free mdirect ("stream of consciousness"), indirect, and implicit perspectives (where a character's perspective is indicated m a more "remote way" through the use of verbs of perception, modal verbs, or the use of definite and indefinite descriptions). The strongest perspective is that of the direct mode, where the responsibility for content and wording is attributed to the character. The weakest perspective is that of indirect speech and implicit perspectives, where the narrator exerts greater influence over the wording of the utterance or perceived event. They indicate this by assigning V-POINT to both the character's space and the BASE. Their analysis works well for Potawatomi, however, where content spaces are not intruded upon by narrators. 140 So, although we have seen that in many places the V-POESTT of fictive narrator and character are conflated, here is an instance where the separate domain of fictive narrator serves nicely as the locus for internal viewpoint. 7.5 Discussion The mental space configurations given in this section are summarized in (20). The columns represent types of discourse. The first division is by genre: Everyday discourse as opposed to narrative discourse. Within narrative, the information types of foreground and background can be classified as 'external perspective', in contrast with the various types discourse covered by 'internal perspective': Direct speech, vividness, and epistemic distance. The rows of the table indicate the location of BASE, V-POINT, and FOCUS CONTENT^ which are given with reference to a domain of spaces; either "Reality" (R), Narrative (N) or Character (C). FOCUS CONTEXT i^ indicated by presence ("Yes") or absence ("No"), and if present, whether the FOCUS is on the Speaker or Hearer. The bottom row of the table represents the sentence pattern used for each type of discourse, either the Conversational Construction (CC) or Narrative Construction (NC): 141 (20) MENTAL SPACE CONFIGURATIONS AND SENTENCE PATTERNS DISCOURSE GENRE EVERYDAY DISCOURSE NARRATIVE DISCOURSE Foreground Background Direct Speech Vividness Epistemic Distance External Perspective Internal Perspective BASE R R R C R R V-POEVT R R R C N N FOCUS CONTENT R N N C N N FOCUS CONTEXT Yes (Speaker or Hearer) No Yes (Hearer) No No Yes (Speaker) Sentence Pattern CC NC CC CC CC CC In everyday speech BASE, V-POINT and FOCUS are all in the "Reality" Domain R. In addition, everyday speech always has a contextual FOCUS on one of the discourse participants (see Chapter 3), and this may shift from the Speaker to the Hearer. We can now differentiate, in mental spaces terms, external and internal viewpoint. With external viewpoint, the V-POINT is outside the FOCUS CONTENT domdan, whereas with internal viewpoint, the V-POINT is inside the FOCUS CONTENT domdan. By this definition, everyday speech has internal perspective. In narrative foreground sentences, BASE and V-POINT remain in R, however FOCUS moves to the Narrative Domain N. Background information shares most of its configuration with the foreground, but differs in having a contextual FOCUS on one of the discourse participants; namely the Hearer. The configuration for reported speech is very similar to that of everyday speech, in that BASE, V-POINT and FOCUS are all within the same domain. The difference is 142 the domain is now shifted to the domain of the character, which becomes a new deictic center. Vividness is represented by V-POINT shifting to the narrative domain, while BASE remains in R. Epistemic distancing shares this configuration, but has a contextual FOCUS on one of the discourse participants, in this case, the Speaker. We now come to the use of the CC and NC, which can now be stated in terms of mental spaces. The only discourse type to use the NC is narrative foreground. If we reasonably take narrative foreground to be representative of the narrative genre (or metonymic for it), the use of the NC in these sentences efficiently distinguishes narrative from everyday speech. A primary function of the conversational and narrative patterns is therefore to indicate which Domain, "Reality" or Narrative, respectively, is in FOCUS. The types of narrative discourse that are represented by the CC all share aspects of their configurations with everyday speech. First, reported speech, vividness and epistemic distance all share internal perspective, or V-POINT inside the Domain that contains FOCUS CONTENT. As noted above, this is also the case with everyday speech. The remaining discourse type to account for is background information, which shares with everyday speech the profiling of a discourse participant. Epistemic distance also profiles a participant (in this case, the speaker), which provides an additional motivation for the use of the CC, besides internal perspective. A primary function of the CC inside narrative is therefore to reference ground by indexing the use of the CC in everyday speech, the prototypical discourse of the "Reality" Domain. 143 7.6 Summary This chapter has presented a Mental Spaces theory analysis that motivates the use of sentential patterns of the NC and CC in narrative. The primary function of the NC is to indicate that the Narrative Domain is in FOCUS, a function enhanced by its use only in foregrounded material. The uses of the CC in narrative are each related in someway to the canonical use of the CC in everyday speech. The similarites which motivates its use in narrative are 1) internal viewpoint, as everyday conversation typically has V-POINT inside the focused "Reality" Domain; and 2) a contextual FOCUS on a discourse participant. In everyday discourse, one participant is always profiled. Narrative generally backgrounds the discourse participants, except in the case of background information, which references the Hearer, and epistemic distance, which references the Speaker. I have also proposed a couple of adaptations to the Mental Spaces theory. First of all, the model of perspective shifts given here revises that of Cutrer (1994). Cutrer analyzes internal viewpoint (such as the use of the historic present) as a BASE shift to a V-POINT within the narrative, either a character, the implied author, or a fictive narrator. I have argued that while internal viewpoint may involve a BASE shift (as indicated by deictic expressions), this is not necessary. In fact, internal viewpoint seems to be a matter of degree, involving minimally a shift in V-POINT, and possibly a BASE shift as well. Analyzing internal viewpoint as a V-POINT shift to the domain in focus provides a contrast with external perspective, where V-POINT is outside of the focused domain. Finally, I have argued that Mental Space structures need to incorporate an elaborated representation of ground. The roles of Speakers and Hearers are necessary to 144 characterize and distinguish certain types of narrative discourse, such as background information and the coding of epistemic distance in internal perspective. In everyday conversation, I have shown that an elaborated representation of ground helps to characterize the difference between illocutionary acts, such as statements and questions (see Chapter 3). Ultimately, if Mental Spaces theory is to handle the complexity of discourse, we need to be able to reference the discourse context. 145 Bibliography Cutrer, Michelle. 1994. Time and Tense in Narratives and Everyday Language, University of California: Ph.D. dissertation. Fauconnier, Gilles. 1984. Espaces Mentaux. Paris: Minuit. Fleischman, S. 1990. Tense and Narrativity. Austin: University of Texas Press. Labov, William. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Longacre, R. E. 1976. 'Mystery' particles and affixes. Proceedings of the Twelfth Regional Meeting of the Chicago Linguistic Society. 468-75. Mushin, liana. 1998. Viewpoint Shifts in Narrative. Discourse and Cognition, ed. by Jean-Pierre Koenig, 323-36. Stanford: CSLI Publications. Sanders, Jose and Redeker, Gisela. 1996. Speech and Thought in Narrative Discourse. Spaces, Worlds, and Grammar, ed. by G. Fauconnier and E. Sweetser. Chicago: The University of Chicago Press. 146 8 Obviation in Potawatomi 8.1 Introduction Obviation is an aspect of Potawatomi grammar worth examining in this study, since, like the use of independents, conjuncts, and the preverb e-, it has different uses in syntax and discourse. In Chapter 10, 1 will argue that these uses are related to each other. The goal of this chapter is to describe obviation in Potawatomi in some detail, since this is an important topic in Algonquian studies, and its use in Potawatomi has not been given much attention in the descriptive literature. Potawatomi also provides an important case study, since its use of obviation places it between such languages as Fox, with significant discourse obviation, and Ottawa, with predominantly syntactic obviation. Based on a detailed study of a traditional narrative, I present a mechanism that would allow a language with discourse obviation to become reanalyzed as a syntactic obviation language, and argue that Potawatomi is an example of this change in progress. 8.2 Background Obviation is a grammatical phenomenon found in Algonquian languages that signals disjoint reference in third persons.^ In a given context, one third person will be designated proximate, and others are marked obviative^ The marking of obviative status ^ Kutenai (a linguistic isolate spoken in British Columbia, Idaho and Montana) also has obviation (involviing first and second as well as third persons) and inverse marking (see Dryer, 1992). Some Algonquianists speculate that Kutenai was a source of diffusion for obviation in Algonquian. ' The earliest use of the term 'obviative' is in Cuoq (1866). 146 occurs on nouns, and is co-indexed by verbal agreement marking. The obviative is the marked category; proximate nominals do not receive special marking on nouns or verbs. Obviation has been compared to switch-reference systems (see Jacobsen, 1967), and within third person, both indicate disjoint reference.'' As Jacobsen points out, though, switch reference relates participants within a narrated event at a local level (across clauses, or adjacent sentences) without reference to the speech context."^ Obviation, on the other hand, also encodes information about the relative status or importance of a referent in a narrative, which indirectly references the speech context, that is, the narrator's ranking of participants. Rhodes (1985) argues against obviation being a property of person marking in part because it is not illocutionary, perhaps in Jesperson's sense of person 'proper' being about distinguishing speech act participants from non-speech act participants (Jesperson, 1924), and also, perhaps, in order to encourage non-Algonquiansts to avoid the use of 'fourth person.'^ This terminology is indeed misleading and confusing, however rather ^ Switch reference systems also indicate coreference, often having paired markers for 'same subject' / 'different subject'. "^ In this sense, switch reference is not deictic, although it is cohesive. Switch reference therefore does not belong to the grammatical category 'person'. One is also less likely to make this claim than for obviation, since the markers of switch reference are generally aspectual suffixes, where as obviation in Algonquian languages is bound up with person/number inflections. ^ The earliest reference to obviation as 'fourth person' seems to have been Uhlenbeck (1909). Algonquian researchers in the 1960's and 70 's commonly used the term: Frantz (1966) for Blackfoot (probably after Uhlenbeck), Rhodes (1976) for Ojibwa, although Wolfart (1973) avoids it. Although the terminology has been abandoned by Algonquianists, it can still be found m general descriptions of obviation, as in Mithun (1990). 147 than avoid treating it as a person, I will continue the practice of the majority of Algonquianists in calling it a distinction within third person.'' In any case, it seems that obviation is at least in part illocutionary, in the sense that within discourse it references the speech context. Several researchers have provided descriptions of obviation in various Central Algonquian languages. Contemporary descriptions include Wolfart (1973), and Dahlstrom (1988) for Cree, Goddard (1984; 1990) and Dahlstrom (1986) for Fox, and Rhodes for Ojibwa, the Ottawa dialect, in particular (1976; 1985; 1990a; 1992; 1993; 1994). Earlier, more limited descriptions of the basic phenomena include Michelson (1921; 1925) for Fox, Bloomfield for Eastern Ojibwa (Bloomfield, 1958), and Hockett for Potawatomi (1939a; 1939b; 1948a-d; 1966). The basic distribution of obviation is as follows: within sentences, there are two contexts for obligatory obviation: third person possessors control obviation of possessees, and when third persons are clausemates, one must be proximate, and the others obviative. There is some control of obviation across clauses, and at least in some languages, across pairs of sentences that have a close semantic relationship. Within discourse, in many languages, obviation is used to mark the relative status of nominals: the higher ranked nominal (usually the "hero" of the discourse) will be marked as proximate, and other third persons will be obviative. "^ Arguments against the use of the term 'fourth person' are mostly made on the basis of negative evidence. Rhodes (1985) bnngs up the pomt that there is no distinction made within the pronoun system that would support a fourth person, and Goddard (1990) notes that "it is either not intended literally or not supported by any morphological or syntactic arguments" (p. 317). 148 The organization of this chapter is as follows. Section 8.3 contains a description of obviative inflection on nouns, demonstratives and verbs. Section 8.4 describes syntactic contexts of obviation. Section 8.5 describes uses of obviation in discourse, using a glossed text Crane Boy which is provided in Appendix C. 8.3 Obviative inflection Obviation is a property of nominals.'' Nouns in Potawatomi bear obviative inflections, and verbs inflect for obviative agreement.^ Both animate and inanimate nominals participate in obviation, however, only animate nouns bear obviative inflection. Both animates and inanimates trigger obviative agreement marking on verbs. The examples below show two sentences with possessed subjects. Possessees with third person possessors are obligatorily obviative, so both subjects are obviative. (1) shows a possessed animate where the obviative inflections are on both the noun and the verb. In (2) the possessed inanimate does not take obviative inflection, but its obviative status is registered in the agreement marker on the verb. (In the free translation, "P" stands for proximate and "O" for obviative.): (1) I je mdadsopon wesme e-byat megwa niw iw jE mEdadEsopon EwEsEme e - bya/e -d megwa niw and ten. years more FCT- come\AI -3C still that.OBV ^ Not in the sense of intrinsic properties, such as (logical) animacy, or plurality, but comparible to number, that is, a deictic property. ^ As Potawatomi is a 'pro-drop' language, referents may be expressed by inflections on the verb as well as NPs. We follow the practice of Rhodes (1990a) in referring to both inflections and NPs as 'nominals'. 149 osen e-yenet. OS -En e - EyE -EnE -d father -OBV FCT- be . in . a . place\AI -OBV -3.C Ten years later, he (P) came back and his (P) father (O) was still there. (2) Mskwane i wbiskewagen mEskwa -EnE -w iw wE- bisEkEwagEn be.red\II -OBV -0 .1 that.INAN 3- clothing His(P) jacket (O) is red. (JT3:63:17) The next two examples show the use of obviative agreement when the subject is not possessed. (3) and (4) show obviative agreement with an AI and II verb, respectively: (3) Bama zhe na gete gigabeyen bama zhE na gEtE gigabe#y -En soon EMPH EMPH for. sure boy -OBV e-nemsenet . e - EnEmOse -EnE -d FCT- walk.off\AI -OBV -3 .C Soon the boy had begun to walk off. (AS. 2. 3. 18) (4) Jak bkwezhgasen wa-mijet zhiw jag bEkwezhEgas -En CH.wi- mij -Ed zhiw all cracker -PL CH.FUT- eat.s.t\TI -3.C there gi-tene . gi- te -EnE -w PST- be. in. a. certain. place -OBI/ -O.I All the crackers for her to eat were there. (JS.4.2.048) 8.3.1 Obviative markers on nouns Obviation is marked on animate nouns with the suffix {En}i.^ This appears as /en/ after consonants as in (5) and /n/ after vowels as in (6). (7) shows its use in marking the obviation of a possessee. ^ This is one of three very similar obviative suffixes, as discussed immediately below. The use of curly brackets around a form m the main text mdicates a morphophonemic representation. 150 (5) ni gwakwadeyen . niw gwakwade#y -En that.OBV grasshopper -OBV 'the grasshopper (obv.) ' (6) ni amon niw amo -En that.OBV bee -OBV (7) niw wmezodanen niw wE- mEzodan -En that.OBV 3- parent -OBV 'the bee (obv.) ' 'his parents (obv.) ' This suffix is the same as the plural inflection on inanimate nouns 10 (8) mzen' egen mEzEn' EgEn book (9) mzen'egnen mEzEn' EgEn-En book -PL 'book ' 'books ' In addition, nouns that inflect for obviation are ambiguous with respect to number. (10) and (1 1) show the grammatically animate noun dabyan 'car' possessed by a first person. In the second example, the possessee is plural, which is indicated on the noun by the use of the animate plural suffix {Eg}. In (12) however, the third person possessor requires that the possessee be obviative, and here number of the possessee is not distinguished: (10) ndodabyan 'my car' nEd-Odabyan 1- car (11) ndodabyanek nEd-Odabyan-Eg 1- car -PL my cars This suggests that on the animacy hierarchy, obviative nominals have a lower animacy status than proximate nominals, similar to the status of inanimates. 151 (12) wdodabyanen wEd-Odabyan-En 3- car -OBV 'his car, his cars ' 8.3.2 Obviative agreement markers on demonstrative pronouns There is also a series of demonstrative pronouns that agree with their head noun in obviative status. There is a proximal, medial, and distal series: (13) Singular (animate) Singular (inanimate) Plural Obviative Proximal ode gode node Medial o(w) i(w) gi(w) ni(w) Distal ago e'i egi eni The proximal and medial obviative forms are related to the singular inanimate forms by the inclusion of {n}, which is transparently similar to the nominal suffix. The medial series is commonly used in texts and functions somewhat like a definite article: (14) o kwe 'the woman ' gi kwek 'the women ' ni kwen 'the women (obv.) ' The indefinite pronoun weye 'someone' is unmarked for obviation, i.e. it does not take obviative marking. However, some speakers use weye, a cognate form borrowed from Fox, which in Potawatomi has an obviative form weyeyen. The obviative form is uncommon; it shows up only once in the corpus, in the text discussed later in this chapter. 8.3.3 Obviative agreement markers on verbs There are three different obviative agreement markers on intransitive verbs, {En}i, {En}2, and {EnE}. These suffixes were historically three different suffixes *-ali, *-ilii, and *-eli- / *-ili-2 (the cognate suffixes occur as three different suffixes, -an -in 152 and -ini respectively in the Ottawa dialect of Ojibwa)/^ Because Potawatomi has merged short -a and i to schwa, the two are now homonymous, and the difference between them and the third is slight. Given two related morphophonemic processes involving schwa — insertion between consonants and syncope — the fact that there are three different suffixes is easily overlooked. The following briefly outlines the evidence demonstrating their synchronic distinctness: {-En}i from *-ali: Sequences of *wa contract to lol (historically short o, but short and long o have merged in Potawatomi). So verbs in {-shEnw} 'stand, lie, fall' end in /-on/ in the obviative, as in the independent verb wjeshnon 'he (OBV) lies beneath', which is morphophonemically ( Oi EshEn w -En | . {-En} 2 from *-ilii *i induces palatalization of a preceding consonant. This suffix is found on the obviative form of the AI participle, as in majinjen 'he (OBV) who leaves', which is morphophonemically {maji -EnE -d -En | {-EnE} from *-eli- / *-ili-2: ^^ The final obviative marker in this hst is either *-eh- or *-ih-. Fox, which would provide the necessary evidence for deciding between them, is ambiguous with respect to these two forms. Also note that *-ilii, and *-eli- / *-ili2- occupy different positional slots, *-eli- / *-ili2- occurring inside of -ilii. 153 The attestation of this suffix is found in Independent II verbs, where the final schwa is retained due to a deleted final glide -w as in wangoyane 'it (OBV) is a hole' which is morphophonemically {wanEgoya -EnE -w} . In many cases, the form of the obviative suffix is ambiguous in surface forms. For example, without respect to the historical origins of the suffixes, one might analyze the sequence /net/ in e-nemsenet 'he/she walked off as {En}i or {En}2 followed by {d} (devoiced) with a connective schwa {E} inserted between them, or even {EnE} followed by {d}. As one can see by my morphemic glosses throughout this chapter, I assume a somewhat abstract analysis: that the use of the obviative suffixes in Potawatomi are consistent with the use of the cognate suffixes in Ottawa, as described by Bloomfield (1958) and Rhodes (1976). {-En}i besides agreement on nouns, is also used to mark obviation of an animate absolutive in the Independent paradigm. Example (15) shows an obviative subject of an intranstive (AI) verb, and (16) shows an obviative object of a transitive (TA) verb. In both cases, the weak vowel {E} in the suffix is deleted following a strong vowel {i} or {a}: (15) majin 'he (O) leaves' maji -En leave -OBV (16) wgi-wabman 'he (P) SOW him (O) ' wE-gi- wabEm -a -En 3- PST- see.s.o.\TA -DIR -OBV When there are two third persons within a clause, as in this example, one must be obviative. When a direct form (here glossed as -DIR) is used, the subject must be proximate and the object obviative. In his tabulation of the TA paradigm, Hockett (1948c 154 p. 142) lists forms with first and second person subjects and obviative agreement with a third person object. These are ungrammatical for modem day speakers, as they are for Ottawa speakers as well (Rhodes, 1976 p. 204). Since they do not show up anywhere in the corpus, I suspect that they may have been ungrammatical for Potawatomi speakers in the 1940'saswell. The suffix{En}2 is used to mark obviative agreement on participles. AI participles take two markers of obviation, the first one shows agreement as an obviative with respect to the participle itself ('internal obviation'), and the second obviative suffix indicates obviation with respect to the rest of the sentence ('external obviation'). The first marker is {EnE} and the second marker is {En}2, which induces palatalization on the preceding consonant: (17) ni amon zazbakdokenj en niw amo -n CH . zizEbakwEdOke -EnE -d -En that.OBV bee -OBV CH .make . sugar\AI -OBV -C -OBV.l the bees who were making honey (AS. 2. 2. 032) The suffix{-EnE} is the most common obviative suffix; it shows up in both independent and conjunct inflections. In transitive verbs, there are obviative inflections in the Independent paradigm. These, as noted above, use {En}i: (18) wwabman 'he (P) sees him (O) ' wE-wabEm -a -En 3- see.s.o.\TA -DIR -OBV.l (19) wwabmegon 'he (O) sees him (P) ' wE-wabEm -EgO -En 3- see.s.o.\TA -INV -OBV.l In the Conjunct paradigm, however, obviation is marked on transitive verbs solely by the use of theme signs: 155 (20) wabmat 'he (P) sees him (O) wabEm -a -t see.s.o.\TA -DIR -3 . C (21) wabmegot 'he (O) sees him (P) wabEm -EgO -t see.s.o.\TA -INV -3 . C 8.3.4 Obviative agreement in participles Participles agree with their head noun in obviative status, as shown by the following examples: (22) Ngodek me se gwakwade e-ndo-mdagwayet nEgOd -Eg mE sE gw akwade e- nEdo- mEdagwayE -d one -LOG EMPH EMPH grasshopper FCT- try- have . f un\AI -3C e-yabtenibek, e-bme-nkweshkwat e- YabEtEnibEn -g e- bEmE- nEkweshkEw -ad FCT- midsummer -O.C FCT- along- meet.s.o.\TA -3/3 '.C bee -OBV [amon zazbakdokenj en] amo -n CH . zizEbakwEdOke -EnE -En CH.make. sugar\AI -3'.P -OBV -OBV.P Once a grasshopper was going along, having fun in the middle of summer, and he (P) met [bees (O) who were making honey]. (AS:2:2:001) (23) wgi-gkeno 'mowan [neshnaben wE- gi- gEkEno'EmEw -a -n EnEshEnabe -n 3- PST- teach. s.o\TA -DIR -OBV person -OBV wa-zhi ' enet] . CH.wi- Ezhi' -EnE -d CH.FUT- be.in.a.certain.place?\AI -OBV -3.C ... he (P) taught [the people (O) who (O) were there]. (JS.4.3.029) (24) wgi-sawan [niw wE- gi- sa -wan niw 3- PST- put . s . o. in. a. certain. place\TA -35/3'. I that. OBV gokoshesen ga-gizswawa j en] gokosh -es -En CH.gi- gizEswa -wa -d -En pig -DIM -OBV CH.PST- cook.s.o.\TA -PL -3C -OBV.P ...he set (P) out the roast pig (O) (.TS.4.2.40) 156 8.3.5 Second Obviative Hockett, (1939b; 1966) describes the use of a second obviative in Potawatomi. The example he gives is shown in (25) which has a doubly possessed nominal. Both nominals are obligatorily obviative, because they are possessed by third persons. The first possessee, okmesen 'his grandmother' takes one obviative marker, and the second possessee dennimnen 'her husband' takes two in succession: (25) nos okmesen dennimnen n-#os #okEmEs -En wEdE-EnEnE -Em -En -En 1- father grandmother -OBV 3- man -POSS -OBV -OBV my father's grandmother's husband He notes however, that in most contexts, two non-coreferent obviatives receive only single obviative marking, as in the following example, where the tree, fellow raccoons, and man are all marked as obviative: (26) Iw je o esben e-gdegozit iw jE ow esEbEn e- EgEdEgOzi -d and that. AN raccoon FCT- climb. up\AI -3.C neko mtegwen, wich- esbenen nEko mEtEgttO -En w#ij- esEbEn -En used. to tree -OBV 3. fellow raccoon -OBV e-mkewat, e-niswebnemwat e - mEkEw -ad e- nisEwebEnEmEw -ad FCT- find.s.o.\TA -3/3'. C FCT- throw. down . to . s . o . \TA -3/3'. C niw neshnaben, neko e-nsat niw EnEshEnabe -n nEko e- nEs -ad that. OBV man -OBl^used.to FCT- kill.s.o.\TA -3/3'. C o neshnabe. ow EnEshEnabe that. AN man The raccoon (P) would chmb a tree (O), find his (O) fellow raccoons (O), and throw them (O) down to the man (O); and the man (P) would kill them (O). (HO.005) 157 Here is another example where there are three referents, and one of the referents is possessed. Neither of the two obviatives, however, is inflected as a second obviative: (27) Ni je wgyeywan gi ni jE wE- #gye#y -wa -En giw and so 3- mother -35.POSS -OBV those. AN gigabesek gi-majingon gigabe#y -s -Eg gi- majin -EgO -En boy -DIM -PL PST- take . s . o . away\TA -INV -OBV.l nenwen . EnEnE#w -En man -OBV 'And SO a man (O) had taken away the boys' (P) mother (O). ' (JS. 4. 1.002) Hockett (1966) remarks that second obviative forms are "rare, and perhaps avoided as 'awkward'" (p. 64). He also notes that that there are no instances where a possessor is obviative and the possessee second obviative (related dialects such as Ottawa have forms for obviative possessees). Second obviative forms and possessee obviatives appear to be no longer in use today, at least, there are no instances in the present corpus. ^^ Since they were falling out of use in the 1940's when the speech community was still quite robust, and since younger speakers of other close dialects like Ottawa (Rhodes, 1993), modern Potawatomi speakers' use of first obviatives only seems to be the completion of this natural change, although attrition as a factor cannot be ruled out. ^' The expected form for an obviative possessee would be the suffix {EnEw}, based on the Ottawa suffix as cited in Bloomfield (1958), so 'his (obv) book' should show up as wde-mzenegne. Hockett's 'second obviative' may m fact be a spurious l-nl, reflecting a strategy used in Ojibwe to avoid final vowel deletion. 158 8.4 Syntactic obviation There are several syntactic domains for the control of obviation. These include two obligatory contexts for obviation: within the phrase (obviation of possessees), and within the clause. Across clauses, subjects of main clauses often control the obviation of subjects in subordinate clauses, however the control here is less strong. One additional context is that of 'sentence clusters'. Each of these contexts is discussed below using data from Potawatomi. 8.4.1 Within phrases At the level of the phrase, the obviation of a noun possessed by an animate third person is obligatory. The example in (28) shows this obligatory obviation when the possessor is third person, whereas in (29) and (30), first and second person nominals do not trigger the obviation of the possessee. (28) wmeshomsen 'his grandfather' wE- mEshomEs -En 3- grandfather -OBV (29) nmeshomes 'my grandfather ' nE- mEshomEs 1- grandfather (30) gmes homes 'your grandfather' gE- mEshomEs 2- grandfather Obviative possessees trigger agreement when they are the subject of intransitive verbs. (3 1) shows this with an animate subject and (32) with an inanimate subject (marked on the verb, but not the NP): 159 (31) Wgwesen me ni gi-ntawen. 3.son=0BV EMPH that . OBV PST-make . a . kill\AI=OBV. I 'His son must have made a kill. ' (JT.3.41.12) (32) Mskwane i wbiskewagen. be. red\II=OEl/. I that.INAN 3. clothing 'His jacket is red. ' (JT.3.63.17) Note that when a possessee is incorporated (in this case, a car), it is not accessible to control: (33) Wgi-bigwdabaneshka o Lucy. 3- PST- have. one ' s . car .break. down\AI . I that. AN Lucy's car broke down. (.TT. 1.44.9) Lastly, conjoined NPs agree in obviative status: (34) Iw je zhe zeshpi e-gi-myanenmat and EMPH a . while . later FCT-PST-dislike . s . o\TA-3/3 ' . C [niw keweziyen mine niw gigabeyen] . that. OBV old.m&n-OBV and that. OBV hoy-OBV Within a short time, she (P) disliked the old man (O) and the boy (O). (.TS. 4.2. 006) 8.4.2 Within Clauses Within clauses, when there is more than one third person, only one may be proximate; others are obviative, as in the following example: (35) Iw je zhe zeshpi e-gi-myanenmat and EMPH a . while . later FCT-PST-dislike . s . o\TA=3/3 '. C niw keweziyen mine niw gigabeyen. that.OBl^ old.man=OBl^ and fhat . OBV hoY=OBV Within a short time, she (P) disliked the old man (O) and the boy (O). (JS. 4.2. 006) 160 Rhodes argues that beyond this statement of distribution, we can say that control of obviation follows the relational hierarchy, where subjects > primary objects > secondary objects^^ > possessors of obliques. The following sentences demonstrate control of obviation in Potawatomi, in accordance with this hierarchy: (3 6) Subj ect > Primary obj ect: I je kezhyep ogeman e-gi-widmowawat and early leader=OBl^ FCT-PST-tell . s . o\TA=35/3 ' . C "Wabozo se wi o ezhcheget." rabbit EMPH EMPH that. AN CH . do . things . a . certain . way\AI=3 . P Early in the morning they (P) told the leader (O), "Rabbit is doing that. " (JS.4. 1.006) (3 7) Subj ect > Primary obj ect (benefactive) : Iw je o nene e-gi-wzhekwat and that. AN man FCT-PST-build. f or . s . o\TA=3/3 ' . C niw keweziyen mine niw gigabeyen that.OBV old.man=OBV and that.OBV boy=OBV wa j -danet . CH . together-live . in . a . certain . place\AI=Oi?V"=3 . C The man (P) built a place for the old man (O) and the boy (O) where they could live together. (JS. 4.2.007) (38) Subj ect > Primary obj ect (di transitive verb) : E-gi-dkobdot wewene FCT-PST-tie. s. t. \TI2=OBJ=3/0 . C carefully e-gi-majidot e-gi-minat FCT-PST-take. s. t. \TI=OBJ=3/0.C FCT-PST-give . to . s . o\TA=3/3 ' .C niw ogeman, "Ode," e-nat. t]nat. OBV leader=OBV this FCT-say . to . s . o\TA=3/3 ' . C He (P) tied it good, took it and gave it to the chief (O) . "Here, " he (P) said to him (O). (.TS. 4. 1.017) '^ For an analysis of primary and secondary objects in Ojibwe see Rhodes (1990b). 161 (39) Primary object > Secondary object (optional for modem-day older speakers): Nbegwzemwa niw gigosen. 1-dry. for. s.o. \TA-DIR. I those . OBV fish-OBV I'm drying those fish for him. (POEX00287) Li sema(n) wgi-minan Biliyen. Lee tobacco=OEl^ 3. PST-give. to. s.o\TA=DIR=OBV. I Billy=OBV Lee (P) gave Billy (O) tobacco ((O)). (MD.245) (40) Subj ect > Possessor of oblique Zhiw wbekwnanek niw there back=PL=LOC that. OBI/ gagtanagoyen [e-] ne-pepegwzot crocodile=OBl/ FCT- start . to-DUP . leap\AI=3 . C e-ne-gwagwashkze ' ot . FCT- start. to-DUP. jump\AI . 3I=3.C So he (P) began to leap and jump there on the backs of the crocodiles (O) . (MD. 1.1. 043) Within this statement of distribution, however, lies some controversy. The disagreement centers on the analysis of inverse verbs. Briefly, Potawatomi (and other Algonquian languages) have verbal morphology which indicates whether the inflections for person/number agreement on some transitive verbs are the properties of the subject or the object. In (41) below, the verbal prefix {nE-} in both (a) and (b) is an agreement marker for first person. The direct suffix {-a} in (a) indicates that the prefix agrees with the subject, and the inverse suffix {-EgO} in (b) indicates the prefix agrees with the object. (41) a) Ngi-wabma o Njan. nE-gi- wabEm -a ow njan 1- PST-see\TA -DIR.I that. AN John 7 saw John. ' 162 b) Ngi-wabmek o Njan. nE-gi- wabEm -EgO ow njan 1- PST-see\TA -INV.I that. AN John 'John saw me. ' With respect to obviation, the difficulty lies with examples like (42b), where the object appears to control the obviation of the subject (Rhodes, 1993): (42) a) Wgi-wabman wgwesen. wE-gi- wabEm -a -En w#gwEs-En 3- PST- see.s.o.\TA -DIR-OBV.I 3-son -OBV 'Hei (P) saw hisi sorij (O). '^^ b) Wwabmegon wgwesen. wE-gi- wabEm -EgO -En w#gwEs-En 3- PST- see.s.o.\TA -INV -OBV. I 3-son -OBV 'His, sorij (O) saw hinii (P). ' Some (Anderson, 1992; Dahlstrom, 1988), maintain that direct verbs and inverse verbs have the same syntax, and the difference is a matter of morphology. When it comes to obviation, these analyses are limited to a general statement of distribution as first given above. Rhodes (1990a; 1994), however, argues that inverse verbs have their surface grammatical relations reversed from their 'notional' grammatical relations. One of the benefits of this analysis is it allows one to make a broader statement about the distribution of proximates and obviatives within clauses, as determined by the relational hierarchy, so long as the ranking based on surface grammatical relations. I will return to these different treatments of inversion in Chapter 9. ^'' Rhodes (1 993) argues that these examples illustrate the Possessor Constraint, the goal of which is to avoid of conflicts in obviative marking. This constraint rules out as ungrammatical cases those readings m which a subject would be possessed by an object, or where a primary object would be possessed by a secondary object 163 8.4.3 Sentential (cross-clausal) obviation In cross-clausal obviation, a third person subject of a main clause controls the obviation of a third person subject in a subordinate clause. In general, this holds for complement, relative, and adverbial clauses: (43) Complement clause, (a) and (b) show 'copying to object' where the verb in the main clause inflects for the subject of the subordinate clause. So (a) would more literally read 'he-saw-him a squirrel running along'. In (c) there is a logical relation of subordination, however the second clause is grammatically an adjunct. Note in all three cases, the lower clause verb inflects for agreement with its obviative subject: (a) Bama zhe na mine e-wabmat (kwekseyen later EMPH EMPH again FCT-see . s . o\TA=3/3 ' . C squirrel=OBl^ e-bmebtonet] . FCT-run . along\AI = OBV"=3 . C Later on, he (P) saw a squirrel (O) running along. (AS:2:2:021) (b) I je o gigabe e-gi-nsaknek and that. AN boy FCT-PST-open . s . t . \TI = 3/0 . C e-gi-mkowat [niw ndemozeyen zhiw FCT-PST-find\TA=3/3 ' . C t]nat . OBV old. woman=OBl^ there e- j ibdebnet] FCT-sit\AIO=OBV"=3 . C So the boy (P) opened it and found the old lady (O) sitting in there... (JT:4: 2:046) (c) Wika zhe e-gi-bige-yekzet o wizhok ever EMPH FCT-PST-tired-be . tired\AI=3 . C that. AN whale [zhiw pene e-chikaznet niw gigabeyen] . there always FCT-play . a . game\AI=0BV=3 . C that.OBV boy=OBV The whale (P) got tired of the boy (O) always playing there. (AS:2: 1:020) 164 (44) Adverbial locative clauses (a) licensed by a relative root, animate subject I je wsezeyma e-zhyat and 3 . older . brother FCT-go . there\AI=3 . C [e j e-nim' edinet] . CH . where-dance\AI=0BV=3 . C So the older brother (P) would go to dances [where they (O) dance]. (JS.4.2.003) (b) adjunct, animate subject I je e-byat [ibe angonoyen e j e-odankwenet] and FCT-come\AI = 3 . C there ant=OBV CH . where-have . a . town\AI = OBl^=3 . C When he (P) got to the ant hill... [where they (O) have a town] ( JS:4:1:013) (c) . . . e-gi-ma j inat niw ndemozeyen FCT-PST-take. s.o.away\TA=3/3 ' .C that . OBV old. woman=OBV [ibe wigwamek ga-je-yenet] . there house=LOC CH . PST-where-be . in . a . place\AI=OBV"=3 . C ...he took the old lady [to the house where she (O) stayed]. (.TS.4.2.068) (d) Ode gigabe e-gi-ma jit e-gi-byat this boy FCT-PST-leave\AI=3 . C FCT-PST-come\AI=3 . C [odanek neshnaben eyenet] . . . town=LOC Indian=OBV CH . be . in . a . place\AI = OBV"=3 . C This boy left and came [to where there was an Indian village]. More literally: he (P) left / he (P) came to a town /Indians (O) were there (.TS:4:3:029) (e) adjunct; inanimate subject Bama zhe na e-byawat [wigwam soon EMPH EMPH FCT-come\AI=35 . C house ga-tenek] . . . CH. PST-be. in. a. certain . place=OBV=0 . C Soon they came to where the house (O) was... (AS:2:3:022) 165 (f) adjunct, inanimate subject Ibe zhe na ga-wje-byat there EMPH EMPH CH . PST-where-come\AI=3 . P e-zhe-gche-ma j it e-byat ibe FCT-in. a. certain. way-really-leave\AI=3 . C FCT-come\AI=3C there jajibdebet e-ne-wabet bzhe ibe sit\AI=3.C FCT-start. to-see\AI=3.C EMPH there e j e-gdegankodnek . CH. there-be. spotted, clouds \ I I=OBV"=0 . P He ran to the place where he had come from, and when he arrived, he sat down and he (P) began to see spotted clouds (O) there! [AS: 1:3: 101) (45) Temporal clause: Iw je i ga-nakwneget and that.INAN CH . PST-plan . things\AI=3 . P e-wi-debmat [pi bwamshe FCT-FUT-grab. s.o\TA=3/3 'C then before gwabtonet] . run . to . shore\AI-0BV-3 . C The one (P) that planned it would grab him (O) [before he (O) reached land] . (MD: 1:1:046) (46) Manner clause: "Jo wika weye gkenmasi e-mbot not never someone know. s . o\TA=DIR= NEG.I FCT-die\AI=3 . C [e-wi- j igwgadenet] ." FCT-FUT-lif t . one ' s . leg\AI = OBV"=3 . C "No one was ever known to die with his legs sticking up. " (.TS:4: 1 :030) In general, cross-clausal control of obviation is much weaker than within the phrase or clause. The following examples show cases where clausal obviation fails to hold. In general, temporal clauses referring to time of year as in (47) are not obviative. These types of clauses always have inanimate subjects. 166 (47) Temporal clause: I me se ngodek jejakok that.INAN EMPH EMPH one-LOC crane-PL e-gche-wzhenwiwat e-nme-dgwagek FCT-really-get . ready\AI-35 . C FCT-getting . to . be-be . autumn-0 . C wech-gzhatek CH . towards -be . hot . weather\II-0 . P e-we-bbonshewat FCT-go . and- spend. the .winter. in. a. certain. place- 35. C Once when it was getting close to Autumn, cranes were preparing for spending the winter in the south... (AS. 1.3. 001) It is also possible for the subjects of complement clauses not to be controlled by obviation. This may be more likely to happen if the complement clause subject is highly topical in the discourse. In (48), for example, wegwendek 'somebody' turns out to be Rabbit, the 'hero' of the narrative. (48) Complement clause: Iw je nish wshkabewsen e-gi-nokanawat and two helper=OBV FCT-PST-have . s . o . do . s . t . \TA-35/3 ' . C e-wi-kewabmawat [wegwendek o FCT-FUT-watch.out. for. s.o. \TA-35/3 ' .C whomever -DUB that. AN ezhcheget] . CH . do . things . a . certain . way\AI-3 . P So they (P) had two scouts (O) watch out for [whomever (P) might be doing that]. (JS:4: 1:002) 8.4.4 Sentence dusters The last type of syntactic context for obviation is what Rhodes (1990a) refers to as 'sentence clusters'. In such cases, adjacent sentences "encode a few very specific semantic relationships, viz. temporal proximity, immediate cause-effect, paraphrase, and a few others" (p. 109). I have found what appear to be analogous constructions in 167 Potawatomi, although Hockett punctuates them as single sentences/^ In (49), the clauses are linked by temporal proximity. Just as the old woman approaches the lake, the boy begins to walk off. The third person pronominal subject of the first clause referring to an old woman controls the obviation of the subject of the second clause gigabeyen 'boy (OBV)'. (49) Ibe e-byat j ik-gchegem; bama zhe there FCT-come\AI=3 . C next . to-big . lake soon EMPH na gete gigabeyen e-nemsenet. EMPH for. sure boy=OBV FCT-walk . of f \AI = OBV"=3 . C She (P) came there to the big lake; soon the boy (O) had started to walk off. (AS.2.3.018) In (50) the first clause provides an example of general behavior, referred to in the second clause. The pronominal subject in the main clause controls obviation of the object, and also of the subject of the second clause. (50) E-wabmawat kojesen e-bshkobnanet; FCT-see. s.o\TA=35/3 ' .C bean=OBV FCT-pull . out . s . o . \TA=3/3 ' . C jak zhe na e-zhechgenet . all EMPH FCT-do. things, a. certain. way\AI=OBV"=3.C They (P) saw him (O) pulling out beans (O); he (O) was doing all kinds of things. (JS.4. 1.004) 8.5 Discourse obviation Apart from the restrictions on obviation as noted above, particularly as generated by obligatory contexts such as possessee and clausemate obviation, there is choice ^^ Hockett was very particular about his use of punctuation, and in my translations, I have nearly always preserved his sentence punctuation, although have added semicolons within sentences where I have felt the need to mark a clause boundary. 168 involved in designating the obviation status of nominals. That is, whenever a clause has more than one third person, there is the choice of which nominal to make proximate, and which others will therefore be obviative. For example, given a narrative about two characters, a raccoon and a wolf, it would be grammatical to say either of the following: (51) a) The raccoon (PROX) saw the wolf (OBV). b) The wolf (OBV) saw the raccoon (PROX). Which form will be used depends on whether the language makes use of discourse obviation. A language with discourse obviation will use it for reference tracking, maintenance of a default ranking of characters to highlight the actions of a "hero", and for narrative-internal viewpoint (this is done with a temporary reordering of the default ranking known as a 'proximate shift') (Dahlstrom, 1988; Goddard, 1984; Goddard, 1990).'^ So given a language (or dialect) with discourse obviation, the expected obviation status of the two nominals would be as in (51a) if the raccoon is the main character in the narrative. If the speaker uses (5 lb) where the main character is obviative, we would expect to find some kind of focus on the secondary wolf character which prompts the status shift. While Central Algonquian languages in general have syntactic obviation, not all make significant use of obviation for discourse/stylistic purposes. Rhodes (1985) points out that while some languages maintain proximates for large stretches of discourse (known as "proximate spans"), others have spans approximately equal to a sentence. Examples are reproduced below of Fox, which has discourse-level spans, and Plains ^"^ Internal viewpoint is used to shift focus to a character, or to represent the narrative as coming from a particular character's point-of-view. (See Chapter 6 for a discussion.) 169 Cree, which has sentence-level spans. In the Fox example all of the proximate references refer to a man, and the obviative references refer to his son. In the second sentence, the son remains obviative, even though he is there is nothing in this sentence to induce syntactic obviation. (52) Fox, cited in Dahlstrom (1996): i>nina>h=ca>hi='pi e>haskacipwi>ha>ci, mese>='nah=meko peno>ci e'h=isihkawe>nici, i>ya>h e>h=oci-pemi-kohkihkawe>nici. "i>ya>h=ca>h=ye>hapa ki>si-pye>hapa!" e>h=isite>he>ci. So then, it is said, he (P) got tired of waiting for him (O), and he (P) followed his (P) son 's (O) trail leading away. His (O) footprints led pretty far away, and over there his (O) footprints turned back and continued on. "He must have gotten back already! " he (P) thought In this Plains Cree example, the proximate referent is reset for each sentence. In the third sentence, the proximate referent resets at the clause level. (53) Plains Cree, cited in Rhodes (1985) Ekwa maci-nikamow. Nimihitowak ekwa sisipak; maka pasakwapiwak. Ekwa pasikow Wisahkecahk; ati-nipahew 6hi sisipa, e-ati-tahkamat oscikwanisiyihk. Kekac e-mescihat, peyak awa apisisisiw napate piko pasakwapiw. Wapamew. (PCT 44:283, 46) 'So hci (P) began to sing. So the ducksj (P) danced, but theyj (P) had their eyes closed. So Wisahkechahk (P) got up, went and killed those ducks (O) by stabbing themj (O) in their j (O) little heads. When hcj (P) was almost done with themj (O), one little onck (P) opened one eye. Hck (P) saw himt (O). Dahlstrom (1988) describes narrative uses of obviation in Plains Cree using a glossed example text, and argues convincingly for some discourse uses for internal viewpoint. However, Plains Cree spans are decidedly short, so by looking through any given text, it is easy to find examples where proximate shifts happen relatively quickly. I suspect that this is such an example. Rhodes (1985), however, also gives Ottawa as a 170 language that has only sentence-level spans, and here it seems to be more clearly the case: (54) Ottawa, cited in Rhodes (1985) Bezhig nini gii-mkadekegban. Aw kiwenzii gii-zhitood wiigwaamens waa-dzhi-mkadekenid niw wgwisan. Gaa-giizhtood dash mii gii-webi- mkadeked aw shkinwe. Pane biindig gii-yaa, gye go gii-wezhho gkizhe wmaanwaang. Niibna dsogon gii-yaa maa wiigwaamensing, gii- baabiidood iw gegoo ji-naabndang. Endso-ggizheb dash gii-zhaa maaba kiwenzii ko gii-ggwejmaad niw wgwisan nmanj iidig gaa- naabndamnigwen. Wgii-gnahmawaan niw wgwisan gaa wii nkwetwaasik niw bi-ggwejmigod mandaagninwan iw ji zhwenmigod. (EO 3 1 : 1-6). [Long ago] a mani (P) fasted. An old man, (P) having built a little hut where hisj (P) soni (O) would fast. After hci (P) got ready, the young mani (P) started to fast. Hci (P) went into the hut, and painted hisi (P) cheeks with charcoal. Hci (P) spent many days in the hut, waiting to see something. Every morning, the old man, (P) came to ask hisj (P) soni (O) if hci (O) had seen anything. Hcj (P) warned his (P) soni (O) not to answer the well-dressed mank (O) coming to ask himi (O) if hck (O) might bless himi (O). If Ottawa represents one end of the spectrum, with only sentence-level obviation, and Fox the other, with copious use of obviation in discourse, then languages like Plains Cree seem to occupy a middle ground. As will be shown below, Potawatomi also occupies this middle ground. Obviation in Potawatomi is decidedly syntactic, with spans approximately equal to the sentence. However some narrators make limited use of discourse obviation, with clear efforts made at maintaining proximates, and some legitimate cases of proximate shifts. I will demonstrate the difference by first briefly examining a Potawatomi text with syntactic obviation. Raccoon and Wolf in Section 8.5.1. In Section 8.5.2, 1 examine in detail a text with more complex obviation. Crane Boy (the full glossed text of Crane Boy is provided in Appendix C). 171 The analysis of Crane Boy shows that while transitive verbs reflect the use of discourse obviation, intransitive verbs follow the syntactic discourse pattern, and generally have proximate subjects. I argue that a possible bridge between the transitive and intransitive uses of obviation are quote frames (see Section 6.3.2), where intransitive verbs of speech that bracket the direct speech of characters nearly always have a proximate third person subject. Because quote frames in Potawatomi are frequently used to register internal viewpoint (see Section 6.3 for a discussion of internal viewpoint), it seems they have become grammaticalized proximate shifts. I will argue that such cases can provide the means of reanalysis of obviation from discourse-level uses, to obviation only at the level of the sentence and below. 8.5.1 Raccoon and Wolf, a text with syntactic obviation Example (55) below comes from a text that has only syntactic obviation. (This example is reproduced from Chapter 6, example 29; the glossed version can be found in Appendix B.). In each sentence, the proximate nominal shifts so that the subject of the transitive verb of speech is always proximate, and the primary object is always obviative: (55) 5 Gete zhena e-gi-nkweshkwat mwen. Sure enough, he (Raccoon, O) met Wolf (P). 6 "Nshi, gde-ton ne gego wa-mijyan ?" "Brother, do you have anything to e-nat espenen. eat?" he (P) said to the Raccoon (O). 7 "Jo zhe kwech bkeji nde-ton wa-mijyan "Not much, I just have a Uttle toeat nawkwek , " e-nat espen . for my own dinner," the Raccoon (P). said to him (O). 8 Mwe e-natewat , "Wegni je etoyen ? " Wolf (P) asked him (O), "What do you have ?" 172 9 Espen e-nat , "Mteno zhe na bkeji Raccoon (P) said to him (O), "I have gokosh-wzhey ndesa , " e-nat . just a httle meat-rmd," he (P) said to him (O). 10 Mwe e-nat , "Mojma shemshen o wzhey." Wolf (P) said to him (O), "Please feed me that rind." 11 I je o espen msach e-gi-minat . So the Raccoon (P) finally gave it to him (O). (AS.4.2) This is the pattern found throughout the text. Main clause intransitive verbs have proximate subjects, and all main clause transitive verbs are direct, with proximate subjects and obviative primary objects.^'' 8.5.2 Crane Boy, a text with discourse obviation The narrative discussed in this section. Crane Boy, was told by the wife of the narrator in (55) above (the glossed text is provided in Appendix C). While this narrative shares the same syntactic obviation pattern in main clause intransitive verbs as Raccoon and Wolf, the treatment of main clause transitive verbs is very different, following the principles of discourse obviation. In Section 8.5.2.1, 1 examine the discourse obviation features of this text. In Section 8.5.2.2, 1 show that the use of syntactic obviation with intransitives, which are numerically preponderant, tend to mask these discourse obviation features. 8.5.2.1 Discourse obviation features ^' I am specifically referring to mam clauses intransitives here, since subordinate clause intranstives can have obviative subjects by virtue of cross-clausal obviation. 173 Narrative Summary. A summary of the text is as follows: The story begins with cranes preparing for their winter migration. While the adult cranes plan and prepare for their journey, some of their boys begin roughhousing. One boy breaks his arm, and so his parents must leave him behind, provisioned only with one rabbit, fully expecting that he will succumb to the harsh northern winter. After the cranes leave, an old woman hears the Crane-Boy crying and takes pity on him, bringing him to her house to live as her adopted grandson, and to be taken care of until the boy's parents return. The old woman takes care of another boy, but he talks back and misbehaves, abusing her benefaction. After an incident, Crane-Boy evicts him. In the next episode of the story, the Crane Boy rids the old woman of a pesky big wooden spoon that steals their food. Spring returns, and the boy watches for the cranes. Soon they return and Crane-Boy's parents find their son and are overjoyed that he is still alive. Ranking ofnominals. Several researchers have argued that discourse obviation is determined by rankings of participants in a narrative, and suggest rankings for the discourses they analyze (Dahlstrom, 1988; Dahlstrom, 1996; Goddard, 1984; Goddard, 1990; Rhodes, 1985).^^ So, based on the summary given above, I will assume a ranking of participants as follows (using a cinematic metaphor of stars, leads, supporting cast, and 19\ extras ): ^^ Aissen (1997) chooses not to analyze the ranking of referents in discourse, which she says "is a psychological or cognitive task, not a linguistic one, though some of our best information about this ranking may come from Imguistic evidence" (p. 710). As linguistic evidence for cognitive constructs forms the basis of this study, we believe this ranking to be well worth examining from a linguistic perspective. ^' The fact that I have a ready metaphor to hand demonstrates that participant rankmgs are natural for narration, and show up for narratives told using other types of media. 174 Starring role: Crane-Boy. Crane-Boy is the "hero" of the story, that is, the character with whom we empathize the most. He emerges as a character very early in the narrative and remains throughout the rest of the narrative. Much of the narrative is told from his point-of-view. Lead: Old woman. The old woman is introduced shortly after Crane-Boy, and is a character throughout the rest of the narrative. We also strongly empathize with the old lady as Crane-Boy's adoped grandmother and benefactress, although she is somewhat distant and mysterious as well: she seems to have mystical powers (she is something of a culture hero), and for part of the narrative, holds the secret of the curse of the Big Spoon. Supporting cast: Crane-Boy's parents, the Bad Boy, the Big Spoon. These characters occur only in the periphery of the narrative, or else in single episodes. The parents are introduced briefly at the beginning of the narrative and do not re- appear until the very end. The Bad Boy shows up briefly, for part of an episode. The Big Spoon, although certainly a memorable character, also belongs to a single episode. All of these characters, are, in one way or another, the 'bad guys', and serve mainly to highlight the heroics of Crane Boy. 175 Extras: the other cranes, crane children (boys), rabbit (for food). These characters show up only briefly, and are usually in the plural (showing their non- individuation). They are essentially props. This ranking can be summarized as follows: (56) Crane Boy > Old Woman > Crane Boy's Parents > the other cranes Bad Boy crane children Big Spoon rabbit If this ranking bears out, we should expect that much of the time. Crane Boy will be proximate, and that he should rarely be obviative. Characters that are less important, or less central, should be proximates less of the time, and occur more frequently as obviative. And this is the case. If we look at NPs, we find that the most important character, the Crane Boy, gets mentioned as a full NP the most (50 references), and none of these are obviative. Out of 34 references to the old woman, nearly half are obviative (14, and 13 of these are possessee obviatives — which will be explained below). The meager three NP references to Crane Boy's parents are always possessee obviative. In main clause transitive verbs, the ranking in (56) generally holds; the highest ranked nominal on this scale is assigned proximate status. In order the proximate status of highly-ranked nominals, which I will refer to as proximate maintenance strategies. These include the use of possessed NPs, passive verb forms, and inverses. Each of these is discussed below. Possessed NPs. One such device commonly found in this text is the use of possessed NPs. For example, the narrator alternates between referring to the old woman 176 as mdemoze 'old woman' and okmesen 'his grandmother (OBV)'. Okmesen, like many kinship terms and terms for parts of the body, is a dependent noun, which means that is obligatorily possessed. Possessed NPs are obligatorily obviative when the possessor is third person. Since the possessor of 'grandmother' in this text is always a third person, Crane-Boy, okmesen is always obviative. Similarly, Crane-Boy's parents are always refered to as wmezodanen 'his parents (OBV)'. One virtue of using these possessed NPs in a clause with a more topical NP, is they will not interfere with the proximate status of the hero, that is, they do not prompt a proximate shift. In addition, these possessed NPs allow for the maintenance of Crane- Boy' s as the central character in other respects. Consider, for example, that the narrator might have referred to the parent cranes simply as gijejakok 'those cranes' and the Crane-Boy as ni wgweswan 'their son (OBV)'. Yet this is not the case; we are told about the actions of Crane-Boy's grandmother and his parents; not her grandson, or their son. As an interesting comparison, Dahlstrom (1996) finds that for the text she is analyzing, the narrator appears to avoid u^mg possessed NPs, as well as various transitive forms. However, in this case the narrator is trying to maintain multiple proximates (multiple proximates are used when a secondary character shares the status of the main character), so using either a possessed NP or a transitive verb would create obligatory contexts for obviation, and disrupt the dual-proximate status. This means that narrators are not at the mercy of obligatory contexts of obviation, but rather, use obligatory contexts selectively in support of their stylistic goals. Passive verb forms. Passive verb forms are also used in the maintenance of a proximate. Goddard (1990) notes that passives, as well as detransitivized intranstives, 177 are a means of suppressing a potential proximate. The most common passive verb in this text is a speech verb: e-nayek 'he/she was told' . Crane Boy is maintained as a proximate from the end of line 51 to line 56, with three uses of this passive in lines 52, 54 and 56. The point of maintaining Crane Boy as a proximate here seems to be so that we will experience the old woman's reprimands from his perspective. The use of the passive is also noteworthy in line 35, where there is another instance of a reprimand, this time, though, the recipient is the Bad Boy. First there is a proximate shift from the old woman (who is proximate from lines 3 1-35) to the Bad Boy at the end of line 35, where the passive is used. At this point, the grandmother becomes a kind of culture hero, cursing the Bad Boy by turning him into a turtle, inventing the creature we know today. This shifts our focus to the Bad Boy, whose new role is introduced in the next line, in an aside to the listening audience: 'and that's why the turtle (P) doesn't know his parents (O).' Inverse verb forms. A third device used in the maintenance of a proximate is inverse verb forms. According to Dahlstrom (1988), inverses are commonly used "to continue tracking the one salient third person throughout an episode". There are three types of situations where inverses are used in this narrative: when the subject is a possessed NP, when the subject is pronominal, and in sentences with references to the Big Spoon. In the first type, which is the most common, the subject is a possessed NP, which is obligatorily obviated. Since the subject is obviative and the object proximate, an inverse verb must be used. As argued above, these instances represent a particular viewpoint by virtue of the NP that is used, and because they are obviative, do not 178 interfere with maintaining the hero as proximate. The example shown here is uses ni okmesen 'his grandmother'. Other examples with possessed NPs as subjects of inverse verbs occur in lines 6, 8, 43, 58, 77 and 100. (57) Iw se e-yayajmo ' got that.INAN EMPH FCT-tell . stories, to. s.o.\TA= 3V3. C e-bkonyak ni okmesen. FCT-be.night\II=O.C that.OB^ 3 . grandmother=0£y So his grandmother told him stories at night. (AS:1:3:24) The second type has inverses with pronominal subjects. This is less common; there are only three such instances in the text; lines 25, 39 and 42. While it is not immediately clear why the narrator chose to use inverses in lines 25 and 39, we will note that the obviative character is in both cases the old woman, and the proximate, is the expected Crane Boy. Line 42 is discussed in the next section. The third type of inverse occurs when the secondary character, the Big Spoon, is the subject of the sentence. There are seven references to the Big Spoon in the text, five of which have transitive verbs (line 42 with a pronominal reference, and lines 46, 58, 78, and 85). All of these are in the inverse. This makes sense as a proximate-maintenance strategy, considering that the object in all of these sentences is the Crane-Boy, the hero of the story, and the Big Spoon is only a supporting character. The narrators use of references to the Big Spoon has a number of subtleties, which will be discussed next. Big Spoon references. Line 42 is the first reference to the Big Spoon, and he is introduced only as a obviative pronominal, with an inverse verb: (58) I me je wi zhe pene that.INAN EMPH but EMPH EMPH always 179 e-kenongot o gigabe . FCT-talk. to. s.o. \TA=3'/3. C that. AN boy So it must be that he would always talk to the boy. (AS:1:3:42) This type of introduction strikes native speakers of English as odd, as one might expect at least an indefinite NP. ^^ However, as we will see, it is part of a larger strategy to gradually increase the salience of this mysterious character.^^ In the next reference, line 46, the Big Spoon is referred to with an obviative NP as weyeyen 'someone (OBV)'. The use of the obviative form of this indefinite pronoun is very unusual, as usually the proximate form (or the unmarked form weye — see Section 8.3.2 ) is used, even when the agreement infiections on the verb show it to be obviative. Finally in line 51, we find out that this mysterious voice belongs to the Big Spoon. Even though this character is, in fact, very animate, gche-emkwan 'big spoon' is grammatically inanimate, and so obviation is not marked on the NP. The intransitive verb in this clause, however, refiects the fact that the NP subject is in fact proximate: (59) E-nme-gisekwet mdemoze, FCT- get ting .to. be- finish. cooking\AI=3 . C old. woman e-bye-bidgeshkak gche-emkwan. FCT- come- enter . with . body\II=0 . C big- spoon When the old lady is almost through cooking, in comes a big spoon. (AS:1:3:51) '° Interestingly enough, I am no longer a good judge of such things in Potawatomi. I didn't realize the oddness of this sentence until it was pointed out to me by several English speakers who read the translation. ^' Another interesting strategy used by this narrator to create suspense is the interleavmg of episodes. Events involving encounters with the Big Spoon are interwoven with descriptions of the old woman and boy gomg about their routine, and of the boy learning how to anticipate his parents' return. This lasts until the climax of the episode in line 99 where Crane -Boy destroys the Big Spoon. 180 In line 58, the Big Spoon is again referred to as 'someone', but unlike line 46, this time the NP weye is proximate. However, the verbal agreement marker shows it has obviative status: (60) E-nme-zag'ek FCT-in . the . process . of -go . outside\AI2=3 . C e-kenongot mine weye, "Wegni FCT-talk. to. s ■ o. \TA= 3 V3. C again someone what wa-mijyek jejakos?" CH. FUT-eat. s. t\TI=25. P crane=DIM When he went out, again someone spoke to him, "What are you going to eat, little crane?" (AS:l:3:5K) A parallel case can be found in line 85. Line 78 raises the animacy status of the Big Spoon by referring to it as ni nenwen 'that man' : (61) Gete ga-gish-gwap ' ek i mbish, for. sure CH . PST-f inish-scoop . s . t . up\TI=3/0 . C that.INAN water e-nnatagot ni nenwen, FCT-ask. s.o. \TA=3'/3. C that. OBV man=OBV "Wegni wa-mijyek jejakos," what CH. FUT-eat. s.t\TI=25. P crane=DIM e-nayek gigabe . FCT-say. to. s.o. \TA=PASS=3.C boy After he dipped into the water, that man asked him "What are you going to eat, little crane .? " (AS : 1 : 3 : 7 8) In the last reference to the Big Spoon in the narrative (lines 98), the NP iw gche- emkwan 'the big spoon' is used. In direct comparison with line 51 (example (59) above), however, this time the verbal agreement marker is obviative: (62) Bama zhe na e-bye-bidgeshkannek soon EMPH EMPH FCT-come-enter . with . body\II=OBV=0C 181 iw gche-emkwan. that.INAN big-spoon That big spoon came reaching in. (AS : 1 : 3 : 98) The following table summarizes the references to the Big Spoon: Line 42 he (O) Line 46 someone (O), marked as (O) in verbal agreement Line 5 1 big spoon, marked as (P) in verbal agreement Line 58 someone (P), marked as (O) in verbal agreement Line 78 that man (O) Line 85 someone (P), marked as (O) in verbal agreement Line 98 big spoon, maked as (O) in verbal agreement Throughout this episode, the narrator gradually increases the salience of the Big Spoon character in several ways. First, by the type of reference: first pronominal, then indefinite, then full NP. Secondly by choice of NP: in one instance, the Big Spoon is referred to as nene 'man' which is grammatically animate, as opposed to gche-emkwan 'big spoon' which is grammatically inanimate. Lastly, through subtle and clever use of obviation. The first interesting use of obviation is with indefinite pronouns. In line 46, the indefinite pronoun is obviative, as are the ones in lines 58 and 85, although here the obviative inflection on the indeflnite NP is suppressed. This has the effect of making the indefinite seem slightly more like a proximate. Another use is with definite NPs. As an inanimate, gche-emkwan is itself never marked as obviative, its obviative status would only be registered in verbal agreement. In nearly parallel syntactic contexts (an intransitive verb in a main clause), lines 5 1 and 98 show a contrast in the obviative status of the nominal. In line 51, the verb has proximate agreement suffix, but in line 98, the 182 agreement suffix is obviative. Although this may seem counterintuitive, I would argue that this use of an obviative agreement marker in fact increases the salience of the referent: it is 'animate' enough to not only to trigger an obviative agreement on the verb, but to do so even in the absence of another clausal third person that might trigger obviation. This is a logical place for the spoon to have a relatively high salience, since this is the moment when he reaches to steal their food (we clearly view this from the perspective of the people inside the house), and Crane-Boy splits him in two. Proximate shifts. While the ranking given in (56) generally holds for transitive verbs, there are a couple of cases where the Old Woman is proximate, and the Crane Boy is obviative. Such instances, where a secondary character is assigned proximate status, are known as "proximate shifts". According to Goddard (1984), proximate shifts serve to shift our attention or "focus" to a secondary character, or to represent that character's point of view. There are two proximate shifts in the Crane Boy narrative. The first takes place in lines 15-18, when the Old Woman discovers Crane Boy. During this span of sentences, the narrative is told from her perspective. She hears someone crying and approaches the sound. The use of the proximate shift has the effect of adding cinematic vividness, but also represents her epistemic stance as being different from our own (a common effect of narrative-internal viewpoint, as discussed in Chapter 6), since we, the audience know this is Crane-Boy, but the Old Woman does not. The only other example in the text where the default nominal ranking does not hold in a main clause transtive is in line 38, where the Old Woman is proximate, and the Crane Boy is obviative. Since this instance is very short (only one sentence), it is more 183 difficult to say for certain that it has the function of a proximate shift, however there are reasons to think this is the case. This sentence introduces the Big Spoon episode by pointing out that the Old Woman is behaving oddly, telling the boy every day what she will cook for their main meal. Although, at this point, the audience may suspect something strange is going on, we don't find out until later that this is an effect of the Big Spoon 'curse'. This is therefore a likely instance of epistemic distancing, which is a plausible context for the use of a proximate shift. 8.5.2.2 Syntactic obviation features Outside of main clause transitive verbs, this text behaves as if it were only governed by syntactic obviation. Main clause intransitives, for instance, are always proximate. The result is that proximates tend to shift very frequently; if there is a sequence of main clause intransitives with alternating subject referents, proximates will shift every sentence. Because main clause intransitives are numerically preponderant, the overall effect is to mask the discourse obviation behavior of main clause transitives. Most of the rapid proximate shifts that take place accompany the verb of speech e-kedot 'he/she said'. Verbs of speech are very common in Potawatomi narrative, and tend to accompany, or bracket every instance of direct speech (see the discussion of Section 6.3.2). The intranstive verb e-kedot is by far the most common verb of speech used for this purpose. I suspect that the regular use of a proximate subject with e-kedot is a result of grammaticalization of discourse obviation. That proximates would become obligatory in this context makes sense, based on the fact that proximate shifts reflect narrative-internal perspective, and we have already established evidence that verbs of speech are used in 184 Potawatomi to mark narrative-internal perspective (generally the narrator's evaluation of the quoted speech — see Section 6.3.2) Since Potawatomi narratives are frequently are short on description and lengthy on conversation, the rapid shift of proximates in intransitives, particularly intransitive verbs of speech, tends to obscure discourse obviation effects. There are many ways that the discourse ranking of nominals is maintained, as we have seen above, but because of the high frequency of this construction, the opposite may appear to be true. Constructions such as these may act as pivots, paving the way for a language with discourse-level obviation such as Fox, to become reanalyzed as a Cree/Ottawa-type with short spans. Potawatomi seems to be in the process of such a shift. 8.6 Conclusion The goal of this chapter was to describe obviation in Potawatomi, including obviative inflection, as well as syntactic and discourse contexts for its use. While Potawatomi has relatively short proximate spans, I have provided evidence that it has some discourse-level uses of obviation: highly ranked characters tend to be referred to with proximate NPs, speakers use a variety of devices to maintain the hero' s proximate status, and beyond this show subtle control of obviation to represent viewpoint and relative character salience. I have argued that these strategies are largely obscured by the fact that there is an abundance of reported speech in narrative, and that this is a context where obviation has largely grammaticalized to only take proximates. The result is that rapid proximate shifts seem to be characteristic of Potawatomi narrative. It may be that such grammaticalized contexts provide a means of reanalysis of obviation, providing the missing link between 185 languages of the Fox-type, with significant discourse-level uses of obviation, to an Ottawa-type, where the domain of obviation is more strictly syntactic. 186 Bibliography Aissen, Judith. 1997. On the Syntax of Obviation. Language, 73.705-50. Anderson, Stephen R. 1992. A-Morphous Morphology .vol. 62: Cambridge Studies in Linguistics. Cambridge: Cambridge University Press. Bloomfield, Leonard. 1958. Eastern Ojibwa: Grammatical Sketch, Texts and Word List. Ann Arbor: University of Michigan Press. Cuoq, Jean- Andre. 1866. Etudes philologiques sur quelques langues sauvages de I'Amerique. Montreal: Dawson Brothers. Dahlstrom, Amy. 1986. Narrative Structure in a Fox text — . 1988. Plains Cree Morphosyntax, Department of Linguistics, University of California, Berkeley: Ph.D. — . 1996. Narrative Structure in a Fox text, nikotwdsik iskwdhtem, pdskihtepayih! Studies in Honour of H.C. Wolfart, ed. by John D.; Nichols and Arden Ogg. Winnepeg: Algonquian and Iroquoian Linguistics. Dryer, Matthew. 1992. A Comparison of the Obviation Systems of Kutenai and Algonquian. Papers of the Twenty-Third Algonquian Conference., ed. by William Cowan, 119-63. Ottawa: Carl eton University. Frantz, Donald. 1966. Person indexing in Blackfoot. International Journal of American Linguistics, 32.50-58. Goddard, Ives. 1984. The obviative in Fox narrative discourse. Papers of the Fifteenth Algonquian Conference, ed. by William Cowan. Ottawa: Carleton University. — . 1990. Aspects of the Topic Structure of Fox Narratives: Proximate Shifts and the Use of Overt and Inflectional NPs. International Journal of American Linguistics, 56.317-40. Hockett, Charles. 1939a. The Potawatomi Language, Anthropology, Yale University: Ph.D. dissertation. — . 1939b. Potawatomi Syntax. Language, 15.235-48. — . 1948a. Potawatomi I: Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi II: Derivation, Personal Prefixes, and Nouns. International Journal of American Linguistics, XIV.63-73. — . 1948c. Potawatomi III: The Verb Complex. International Journal of American Linguistics, XIV. 139-49. — . 1948d. Potawatomi IV: Particles and Sample Texts. International Journal of American Linguistics, XIV.213-25. — . 1966. What Algonquian is really like. International Journal of American Linguistics, 32.59-73. Jacobsen, William. 1967. Switch-Reference in Hokan-Coahuiltecan. Studies in Southwestern Ethnolinguistics, ed. by Dell H. Hymes, 238-63. The Hague. Jesperson, Otto. 1924. The Philosophy of Grammar. Chicago: University of Chicago Press. 187 Michelson, Truman. 1921. The Owl Sacred Pack of the Fox Indians.vol. Bulletin no. 72: Bureau of American Ethnology. Washington, D.C.: Smithsonian Institution. — . 1925. Accompanying papers.vol. 40: Bureau of Emerican Ethnology Annual Report. Washington, D.C.: Smithsonian Institution. Mithun, Marianne. 1990. The Languages of Native North America. Cambridge: Cambridge University Press. Rhodes, Richard A. 1976. The Morphosyntax of the Central Ojibwa Verb, Linguistics, University of Michigan: Ph.D. — . 1985. Obviation: the Mark of Non-Coreference — . 1990a. Obviation, Inversion, and Topic Rank in Ojibwa. Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society February 16-19, 1990: Special Session on General Topics in American Indian Linguistics, 16S. 101-15. — . 1990b. Ojibwa Secondary Objects. Grammatical Relations: A Cross-Theoretical Perspective, ed. by Katarzyna; Dziwirek, Patrick; Farrell and Errapel Mejias- Bikandi, 401-14: Center for Study of Language and Information. — . 1992. The Syntax of Possessor Obviation in Ojibwe — . 1993. The Possessor Constraint — . 1994. Agency, Inversion, and Thematic Alignment in Ojibwe. Proceedings of the Twentieth Annual Meeting of the Berkeley Linguistics Society February 18-21, 1994: General Session: Dedicated to the Contributions of Charles J. Fillmore, ed. by Susanne; Dolbey Gahl, Andy; Johnson, Christopher, 431-46. Berkeley: Berkeley Linguistics Society. Uhlenbeck, Christianus Cornelius. 1909. Grammatische onderscheidigen in het Algonkinsch. Amsterdam: Akademie van Wetenschappen. Wolfart, H. Christoph. 1973. Plains Cree: A grammatical study. Transactions of the American Philosophical Society, 63. 188 9 The Obviation Construction 9.1 Introduction Like the Independent and Conjunct paradigms and the preverb e-, obviation is another grammatical phenomenon of Potawatomi which has different but related uses in syntax and discourse (described in Chapter 8). These different uses within various Algonquian languages have given rise to two main theories of how obviation works: that obviation is basically a syntactic device, or that it is primarily a function of discourse. It is a descriptive fact that within the Algonquian language family, some languages and dialects regularly employ discourse obviation in narrative while others make little to no use of it. For example, in languages like Fox and Plains Cree, proximate selection is largely determined by the role or status of nominal referents in the narrative, with the most central character, or 'hero', generally assigned proximate status. Proximate spans (where one nominal referent is maintained as a proximate) can last through long stretches of text (Dahlstrom, 1988; 1996; Goddard, 1984; 1990). In Ottawa, on the other hand, the proximate span is equal to roughly a sentence, and proximate selection is based on the grammatical function of a nominal (Rhodes, 1990; 2002). Potawatomi presents an interesting case for these theories, because the use of discourse obviation is not language or dialect specific, but rather appears to depend on the narrator: Jim Spear's texts (such as "Raccoon and Wolf) can be explained solely by reference to grammatical function within a sentence, however Alice Spear's texts (for 187 example, "Crane Boy"), show clear efforts at proximate maintenance/ A satisfactory account of Potawatomi must therefore allow 1) syntactic obviation in the absence of discourse obviation, and 2) some access/use of discourse obviation for those narrators that make use of it. In this chapter, I will argue for a constructional approach to obviation. That is, I will argue that obviation has a very broad function — hierarchically ranking non-coreferent third persons — and this finds different expression across grammatical domains. The advantages of such an approach are 1) it theoretically unifies various instantiations of obviation, 2) helps explain how obviation could be extended to apply in new contexts, and 3) allows us to explain how speakers of one language or dialect may access, to varying degrees, one particular type of obviation, such as its discourse use. In addition, I will incorporate information about Mental Space networks into constructions. Because Mental Spaces theory is designed to handle the representation of viewpoint, it allows us to capture the changes in perspective signalled by proximate shifts. Indexing the Mental Space network inside of constructions allows constructions to "see" what is happening at the discourse level. The format of the chapter is as follows: in Section 9.2, 1 discuss previous analyses of obviation, giving particular attention to a theory I call the "integrated approach" which forms the basis of the present analysis. Section 9.3 presents what I call the "constructional approach" to obviation. Section 9.4 lays out the details of this approach, and discusses how constructions are indexed to Mental Space networks. ^ Grammatical attrition might be suspected here, but is not a Ukely explanation given both were very fluent speakers, narrating texts at at time when the use of Potawatomi was still quite robust. 188 Section 9.5 discusses the relationship between the various obviation constructions, and proposes the concept of "constructional maintenance" to account for the difference between the use of obviation in languages like Fox / Plains Cree, Ottawa and Potawatomi. 9.2 Previous analyses Previous analyses of obviation in Algonquian languages can be grouped as pre-generative (traditional grammatical descriptions), syntax-based, discourse-based, and what I will call the "integrated approach". Each of these is discussed, in turn, below. Pre-generative descriptions of Algonquian languages treat obviation as essentially a discourse phenomenon, where proximates are described as the 'topic' or 'focus' of discourse (Bloomfield, 1962; Hockett, 1966; Wolfart, 1973). A good representation of this perspective is Bloomfield' s description of obviation in Menomini: "The proximate third person represents the topic of discourse, the person nearest the speaker's point of view, or the person earlier spoken of and already known" (1962, p. 38). Later syntactic studies of obviation rejected the notion of proximates being the discourse focus, arguing that this definition of focus is circular (defined only in relation to obviation), and that it does little to explain obligatory contexts of obviation, such as in the case of possession, and clausemate obviation (Dunnigan et al., 1978; Grafstein, 189 1981). While these syntactic studies do not discount the discourse use of obviation, they exclude it from their analyses, as Grafstein states:^ I suspect that one of the reasons for the shortcomings of the traditional approach lies partially in its failure to separate the semantic function of obviation at the level of discourse from its syntactic function at the sentential level. The attempt to describe and predict obviation exclusively in terms of notions such as 'focus' obfuscates the syntactic relationships which hold between proximate and obviative nouns within sentences, (p. 98) While these studies resulted in a much richer description of the syntactic realization of obviation, they were later criticized for disregarding the role of obviation in discourse and its effect on clause and sentence-level syntax (Goddard, 1984). Proponents of discourse-based obviation (Dahlstrom, 1988; 1996; Goddard, 1984; 1990) argue that in any given narrative, the highest ranked nominal referent (the 'hero') will be assigned proximate status, and other nominals will be obviative. This default ranking is sometimes overridden in specific contexts, and the alternation of proximate status is known as a 'proximate shift'. Proximate shifts occur when there is focus on a particular character, or the narrative is presented from a particular character's viewpoint (what we have referred to as 'internal viewpoint' — see Section 6.3). In these cases, the ranking may assign a secondary character proximate status, and other nominals will be marked obviative. An indication that these shifts to secondary characters do not represent the default ranking is that they often require more 'machinery', such as specification with overt NPs (Goddard, 1990). " Aissen (1997), a more recent example of the syntactic approach to obviation, also adopts this tactic: "[t]he ranking of referents according to discourse salience is a psychological or cognitive task, not a Imguistic 190 Inversion is an important part of a discourse-based argument (see Section 2.8 for a description of direct and inverse verb forms). According to this theory, in any given text, direct and inverse verb forms are used to maintain a high-ranking argument as proximate. In any given clause, if a subject is proximate and the object obviative, the verb will be marked as 'direct'. If the subject is obviative and the object is proximate, the verb will be marked 'inverse'. Proximates and obviatives are therefore determined by discourse ranking, and inversion follows from the assignment of obviation status. Richard Rhodes, in several articles, argues against the idea that obviation is discourse-driven, in part because such a theory does not account for languages like Ottawa that do not make significant use of discourse obviation. He argues instead for an integrated theory of obviation that encompasses both syntax and discourse (1976; 1985; 1990; 1992; 1994; 2002). The remainder of this section describes this theory in some detail, because it forms the basis of the present analysis. A summary of the relevant features of this theory is as follows. Within clauses, control of obviation is determined by a hierarchy of grammatical relations, given in (1): (1) subjects > primary objects > secondary objects > possessors of obliques The highest third person on this scale is the 'preferred argument' ("PA"). Within a clause, if anything may be proximate, it will be the preferred argument. The preferred argument then controls obviation of other third persons within the clause, and to some degree, sententially. In languages that have discourse obviation, nominals can be 191 obviated from outside the sentence. In this case, the selection of preferred argument is based on discourse topic: the highest nominal on the topic hierarchy is the preferred argument. Inverse verb forms also play an important part in this theory. It is worth noting at this point that inverse verbs in texts with syntactic obviation occur very infrequently, if at all (a rough estimate would be about 2% of all main clause transitive verbs). ^ While inverses have an obvious function in languages with discourse obviation — that of maintaining a proximate over a span of sentences — their role in syntactic obviation is not apparent (since there is also a passive). If inverses are less important for syntactic obviation, it stands to reason that they should not be common. Whereas in the discourse-central theory direct and inverse verb forms are read off of the mapping of proximate and obviative to surface grammatical functions, for the integrated theory, this mapping takes place between notional and final grammatical relations. That is, with a direct verb, notional and final relations are aligned, but inverse verbs reverse the notional and final relations, and are thus passive-like. In order to lay this out in a little more detail, I have shown the difference between the assignment of the Preferred Argument in a syntactic and discourse obviation language in the tables below. For the purposes of illustration, I have assumed a hypothetical text about Raccoon and Wolf, where Raccoon outranks Wolf on the topic hierarchy, that is. Raccoon is the central character. The example sentences use the transitive verb 'see' which takes a subject and primary object. ^ I have observed this in Potawatomi, and Rhodes (p.c.) notes that this is also the case in Ottawa. 192 The first set of tables show how proximate assignment and inversion would work in a syntactic language. In (2), Rabbit is the final subject, and is assigned preferred argument status (shown in boldface). Using a direct form means that final relations match notional relations. (2) SYNTACTIC LANGUAGE, DIRECT VERB Raccoon saw Wolf Notional Relations Subject > Primary Object Final Relations Subject > Primary Object Preferred Argument PA Characteristic of syntactic obviation languages, if the nominal referents are switched (which is common when characters converse), a syntactic language will generally still assign PA to the final subject, and will use a direct verb, as shown by the alignment of notional and final relations in (3). (3) SYNTACTIC LANGUAGE, DIRECT VERB Wolf saw Raccoon Notional Relations Subject > Primary Object Final Relations Subject > Primary Object Preferred Argument PA Occasionally, however, in syntactic obviation language, an inverse verb form will be used. In this case, the PA is still the final subject, but because the notional and final relations are mismatched, the verb will be inverse, as shown in (4): 193 (4) SYNTACTIC LANGUAGE, INVERSE VERB Wolf saw Raccoon Notional Relations Subject > Primary Object Final Relations Primary Object < Subject Preferred Argument PA In a language with discourse obviation, the PA associates to the highest ranked nominal on the topic hierarchy, in this case, the Raccoon (shown in boldface in (5)). The alignment of notional and final relations means the verb is direct: (5) DISCOURSE LANGUAGE, DIRECT VERB Raccoon saw Wolf Notional Relations Subject > Primary Object Final Relations Subject > Primary Object Preferred Argument PA The association of PA to the topic hierarchy can be illustrated by switching the nominal referents, as shown in (6). Characteristically, a discourse obviation language will now use an inverse verb, which is based on the mismatch of notional and final relations. (6) DISCOURSE LANGUAGE: INVERSE VERB Wolf saw Raccoon Notional Relations Subject > Primary Object Final Primary < Subject 194 Relations Object Preferred Argument PA So while, for a discourse obviation language, obviation is linked to discourse topic, inversion is still a syntactic process, based on the comparison of notional and final grammatical relations. While this approach adds complexity by representing both notional and final relations, it has the benefit of accounting for both syntactic and discourse obviation languages, whereas stand-alone syntax or discourse theories of obviation only account for one type. The analysis of inverse verb forms as a kind of passive operation is a matter of debate among syntacticians studying Algonquian languages, about which I will only add that in the Potawatomi texts I have examined, inverses seem to have a small range of functions; they are mostly used when someone is being scolded, and therefore are probably a device to background the person doing the scolding. If so, this use would be in keeping with passive-like semantics."* 9.3 Constructional approach The approach I will adopt here is basically that of Rhodes, as presented above. The modification I will propose is that obviation is constructional. This means that obviation is essentially a pairing between form (including proximate selection and obviative morphology) and meaning. I will argue that the meaning of the obviation "* For an analysis of inversion as a morphological rather than syntactic operation, see (Dahlstrom, 1 988). This is also the analysis of inverses adopted by Anderson in his discussion of Potawatomi (Anderson, 1992). 195 construction itself is quite broad (ranked non-coreferent third persons), and that it is inherited by constructions that further specify its meaning within particular domains. The family of constructions that inherits obviation illustrates constructional polysemy. That is, constructions, like lexical items, can have multiple semantically related senses that form polysemy networks. In the next section, I will outline how such a constructional approach would work. I will not consider all the details of obviation, which would unnecessarily complicate the line of argumentation. Rather, I will focus on the operation of obviation in a few critical contexts such as possession, clausemate obviation (particularly between subject, primary object and secondary object), sentential obviation, sentence pairs, and discourse. I will also primarily lay out the constructions themselves, rather than the particular spelling of obviative morphology. (The exception will be the case of possession, where having a construction provides a means of distinguishing between the marking of obviation on animate possessees, but not inanimate ones.) 9.4 Obviation constructions in Potawatomi 9.4.1 Possession The smallest domain for the operation of obviation is the phrase, as evidenced by possessed NPs where the possessor is third person. In this case, obviation of the possessee is obligatory. Consider the following example where the possessor is third person and the possessee is a grammatically animate noun, dabyan 'car'. The possessed noun is obligatorily marked obviative with the /-En/ suffix: 196 (7) wdodabyanen wEd- Odabyan -En 3- car -OBV In order to capture the obviation facts with possessed NPs, we will propose the first of several hierarchies, the possession hierarchy, given in (8), where possessors outrank possessees. (8) POSSESSION HIERARCHY: possessor > possessee In general, with regard to such hierarchies, we will say that a nominal that is highly ranked is more likely to be proximate and induce obviation on nominals of lower rank. A first formulation of the Possessee Obviation Construction is shown in (9). This matrix is an abbreviated representation of a construction (or construct, if the information in the matrix is entirely filled in), with information extracted from various parts of the full construction. The matrix includes three types of information, syntactic ("Syn") and semantic ("Sem") and role, or grammatical function information. Divisions of information within these types is represented by horizontal tiers. For the Possessee Obviation Construction, within the role specification, the grammatical function (gf) tier contains the ranking of possessor and possessee, and the person (pers) tier records the person of each nominal. Within the semantic specification, the obviation tier (obv) contains a single value of proximate (PROX +) which is available for linking with the other tiers (only the proximate value is shown in this representation, ^ This abbreviated representation is based on those given in Goldberg (1995), who uses them to link grammatical functions and thematic roles provided by a general construction with the semantic roles provided by individual verbs. "^ The separation of role information from syntactic information is a convention of Construction Grammar, and serves as a means of linking grammatical functions with thematic roles. 197 obviative values (PROX-) will be filled in by default in a later construction, "Default Obviative Assignment, shown in (12)). Every nominal that is also third person is "visible" to the obviation tier, and thus available for linking with proximate. However, the proximate value associates only with the highest ranking (leftmost) nominal on the gf tier that is also third person. The construction will therefore assign proximate status to the highest ranking nominal on the Possession Hierarchy. (9) POSSESSEE OBVIATION (first formulation): Role: gf [ Possessor > Possessee] Syn: pers [ 3 3 ] Sem: obv [ PROX+ ] As stated, the Possessee Obviation Construction has information which will be redundant when we consider the operation of obviation at higher syntactic levels such as the clause and sentence. In order to make a more general obviation construction, we will need to separate out the information that is particular to possession, that is the Possession Hierarchy. The Obviation Construction, given in (10), will then have slots for ranked nominals in an unspecified tier, which will be filled in by particular hierarchies. (10) OBVIATION [ Nomi > Nomj > . . > Nomn] Syn: pers [ 3 3 3 ] Sem: obv [ PROX+ ] 198 The revised Possessee Obviation Construction, in (1 1), contains information inherited from the Obviation Construction, and contributes additional information by specifying the use of the Possession Hierarchy for the gf tier (shown in boldface). '' Given a ranking of specific nominals, then, Possessee Obviation will link proximate with the highest ranked third person nominal.^ ^ This construction includes the grammatical relation ranking; I leave open the question as to whether such rankings are themselves constructional. Richard Rhodes has pointed out (p.c.) that if the hierarchies are constructional, it explains certain gaps in the application of clausemate obviation. ^ Although in this instance, only Obviation is inherited, note that my convention for representing inheritance relationships will be to cite all the inherited constructions, rather than just the immediately inherited parent construction. 199 (1 1) POSSESSEE OBVIATION (revised) TNHF.RTT- ORVTATTON Role: gf [ Possessor > Possessee] Syn: pers [ 3 3 ] Sem: obv [ PROX+ ] The remaining issue to address is the assignment of obviative status (PROX -) to any other third person nominals not associated with proximate status. These values will be filled in by default, as given in (12). The arrows in the construction show that the third person nominals which were not previously assigned to proximate by the inherited Obviation Construction, are now all assigned obviative status. (12) DEFAULT OBVIATIVE ASSIGNMENT Role: gf [ Nomi > Nomj > . . > Nom„ ] Syn: pers [ 3 3 3 ] ^r V Sem: obv [ PROX+ PROX- .. PROX-] To illustrate the full construct wdodabyanen, we will in this instance provide a morphological spelling rule, given in (13) below. This will illustrate that although constructions assign obviative status to both animate and inanimate nominals, only animate ones are given obviative marking. Possessee Obviative Spelling specifies that a 200 grammatically animate, obviative possessee will be marked by the obviative suffix, given here as {-En}: (13) POSSESSEE OBVIATIVE SPELLING syn ^ cat n "^ lex C stem + ^ L word + J '^ prox - J sem anim + syn feat nsuff ] Ixm -En We now return to the example given in (7), of the possessed animate noun wdodabyanen 'his/her car'. The construct wdodabyanen is shown in (14) below. The construct inherits the Possession and Obviation constructions, and these together link the possessor nominal with proximate. The Default Obviation construction supplies the obviative value of the possessee. The noun dabyan is grammatically animate (abbreviated "anim +" in the diagram), and this animacy value unifies with the external semantics to make the construct as a whole grammatically animate. Because the construct is both obviative and animate (as specified in the external syntax and semantics of the construct), Possessor Obviative Spelling applies, supplying the obviative suffix {-En}. 201 (14) wdodabyanen 'his/her car' INHERIT: POSSESSEE OBVIATIVE SPELLING, POSSESSEE OBVIATION, OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT syn r cat n ■N lex f stem + ^ L word + J V. prox ■ ^ sem anim syn ' cat npref ^ role POSSESSOR __ sem ' anim +' prox + Ixm wEd- syn /"cat n lex "A r stem + ^ l^ word - J role POSSESSEE sem r anim + 1 L prox - J Ixm Odabyan syn feat nsuff ] Ixm -En Because Possessee Obviative Spelling only applies to grammatically animate obviative constructs, it will not apply in the case of an inanimate possessee. An example is given in (15) of the inanimate possessed noun wdonagen 'his/her dish'. The possessee is semantically obviative, as specified in the external semantics, but it is not morphologically marked as such: 202 (15) wdonagen 'his/her dish' syn ^ cat n "^ sem fanim 1 lex C stem + ^ ^ word + j V. prox - y syn ' per 3 role POSSESSOR syn i^cat n A lexf stem + "| ^ word + J sem ' anim + role POSSESSEE prox + sem f anim - Ixm wEd- L prox - J Ixm OnagEn 9.4.2 Clausemate The next larger domain to which obviation applies is the clause. Only one third person in a clause may be assigned proximate; any other third persons will be obviative. For syntactic obviation languages, which nominal in a sentence will be proximate is predictable based on its grammatical function. Proximate selection follows the relational hierarchy, where subjects outrank primary objects, and primary objects outrank secondary objects (for a description of the operation of this hierarchy see Section 8.4): (16) RELATIONAL HIERARCHY; SUBJ > P.OBJ > S. OBJ We can then state Clausemate Obviation much as Possessee Obviation, the difference being that Clausemate Obviation inherits the Relational Hierarchy to fill in the values for the ranked nominals (shown in boldface): 203 (17) CLAUSEMATE OBVIATION (first statement) INHERIT: OBVIATION Role: gf [ SUBJ > P. OBJ > S. OBJ ] Syn: pers [ 3 3 3 ] Sem: obv [ PROX+ ] 9.4.2.1 Direct Verb To demonstrate the use of Clausemate Obviation, consider the direct transitive verb in (18), which has a proximate subject and an obviative primary object. In order to make the presentation somewhat easier, we will assign the nominal referent of the subject as RACCOON (esben) and the primary object as WOLF (m 'ewe) (these nominals are not included in the Potawatomi sentence here, but are registered inflectionally on the verb). (18) e-gi-wabmat e - gi- wabEiti -a -d FCT- PST- see.s.o\TA -DIR -3C 'he [raccoon-PROX] saw him [wolf-OBV] ' When the nominal values for subject and primary object are supplied to Clausemate Obviation, the result is the matrix given in (19). The ranked grammatical roles of subject and primary object are supplied by Clausemate Obviation. The Obviation Construction associates PROX+ to the highest ranked nominal on the relational hierarchy, which is the subject. Clausemate Obviation inherits Default Obviative Assignment, which supplies PROX- for the remaining nominal, the primary object. 204 (19) CLAUSEMATE OBVIATION CONSTRUCT, e-gi-wabmat 'he [raccoon-PROX] saw him [wolf-OBV]' INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT Sem: ref [RACCOON WOLF ] Role: gf [ SUBJ > P.OBJ ] Syn: pers [3 3 ] Sem: obv [PROX+ PROX- ] A fully specified construct oi e-gi-wabmat is shown in (20), which shows the information from (19) in its place within the verbal valence. 205 (20) Fully Specified Construct: e-gi-wabmat 'he [raccoon-PROX] saw him [wolf-OBV]' INHERIT: CLAUSEMATE, OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT syn ^ cat V "^ sem C frame SEEING ~^ lex r stem + "^ ^ word - J part 1 #1 [ ] part 2 #2 [ 1 ^ ^ ^' J val r #1 /syn r cat nominal "^ \ per 3 #2 /^syn r cat nominal ^ \ [_ per 3 J -^ sem ''ref raccoon] sem ''ref WOLF^ < anim + anim + > obv prox + obv prox - V. J role gf subj role gfpobj V. ^ L e exp J J L 9 cont J J ^A.1.1 Inverse verb Next we will compare the case of the inverse verb, which will require some refinements of Clausemate Obviation. To accommodate inverses, we will need a construction that specifies an additional tier, which records final relations alongside notional relations (notional relations are represented here as an inverted hierarchy): (21) INVERSE CONSTRUCTION Role: notgf [P.OBJ < SUBJ] Role: fin gf [ SUBJ > P.OBJ ] 206 We will also need to specify that the relational hierarchy in Clausemate Obviation is based on final relations (shown in boldface): (22) CLAUSEMATE OBVIATION (final) INHERIT : OBVIATION Role: fingf [ SUBJ > P. OBJ > S. OBJ] Syn: pers [ 3 3 3 ] Sem: obv [ PROX+ ] Let us now examine the inverse verb e-gi-wabmegot given in (23). This verb differs minimally from e-gi-wabmat (18) in that the final subject of the verb is obviative, and the primary object is proximate. (As above, we will use nominal referents, this time with the obviative WOLF as notional subject, and the proximate RACCOON as notional primary object.) (23) e-gi-wabmegot e - gi- wabEiti -Ego -d FCT- PST- see.s.o\TA -INV -3C 'he [wolf-OBV] saw him [raccoon-PROX] ' When Clausemate Obviation applies, it operates on final relations, where the final subject is RACCOON and the final primary object is WOLF: 207 (24) CLAUSEMATE OBVIATION CONSTRUCT, e-gi-wabmegot 'he [wolf-OBV] saw him [raccoon-PROX]' INHERIT: OBVIATION DEFAULT OBVIATIVE ASSIGNMENT Sem: ref [RACCOON WOLF] Role: fingf [ SUBJ > P.OBJ ] Syn: pers [3 3 ] Sem: obv [PROX+ PROX- ] Clausemate Obviation then inherits Inverse, which applies because the notional and final relations are mismatched (see (25)). It does not change the assignment of proximate and obviative, which were already specified by Clausemate Obviation and Default obviative Assignment. (25) CLAUSEMATE OBVIATION INHERITS INVERSE INHERIT: INVERSION Sem: ref [RACCOON WOLF ] Role: fingf [SUBJ Role: notgf [P.OBJ Syn: pers [ 3 Sem: obv [PROX+ > P.OBJ ] < SUBJ ] 3 ] PROX-] 208 The fully specified verbal construct for e-gi-wabmegot is given in (26), showing the information about final relations (in boldface): (26) CLAUSEMATE OBVIATION INHERITS INVERSE, CONSTRUCT e-gi-wabmegot 'he [wolf-OBV] sawhim [raccoon-PROX]' INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT, INVERSE syn '' cat V "^ sem C frame SEEING "^ lex r stem + "1 ^ word - J parti #1 [] part 2 #2[] V. J val r #1 /syn ' cat nominal " per 3 J #2 /'syn r cat nominal "^ > L per 3 "N sem ^ref WOLF^ sem ^ref RACCOON < anim + anim + > obv prox - ' obv prox + role not gf subj fin gf pobj role not gf pobj fin gf subj V. 6 exp ^ ^ 9 cont y J Because direct and inverse verbs have different inflectional morphology, we will need to create a Direct Construction to parallel the Inverse construction. The Direct and Inverse Constructions can then be inherited by constructions which specify the spelling of direct and inverse morphology (these will not be given here, as discussed above). The Direct Construction is given in (27). The construction states that in a direct verb, final relations are the same as notional relations. 209 (27) DIRECT CONSTRUCTION Role: notgf [ SUBJ > P.OBJ] Role: fingf [SUBJ > P.OBJ] The fully specified verbal construct for (23) can then be restated as follows, which includes the information about final relations (in boldface): (28) CLAUSEMATE OBVIATION INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT, DIRECT syn " cat V ^ sem 1 frame SEEING "^ lex C stem + "] , word - J part 1 #1 [ ] part 2 #2[] ^ ^ ^ J val r #1 /syn ' cat nominal >~~^ \ #2 f^syn f cat nominal ^ \ I per 3 J ~^ per 3 , sem ''ref RACCOON sem ^ref wolf" s < anim + anim + > obv prox + ' obv prox - / role not gf subj fin gf subj role not gf pobj fin gf pobj V. 9 exp ^ ^ ^9 cont ^ J J 9.4.3 Primary Object > Secondary Object Besides notional subjects inducing obviation on notional primary objects, primary objects also induce obviation on secondary objects. In the following sentence, the 210 primary object 'him' is proximate, and induces obviation of the third person secondary object gigosen 'fish': (29) Nbegwzemwa niw gigosen. l.dry.for .3 .o. \TA=DIR. I those=OBV fish=OBV I'm drying those fish (OBV)for him (PROX). [POEX00287] This obviation fact is easily captured using the existing machinery of Clausemate Obviation. Since in this case the subject is first person, it is not visible to obviation. Proximate will associate to the highest available nominal on the hierarchy, which is in this case the primary object. The obviative status of the secondary object can then be filled in by Default Obviative Assignment. (30) PRIMARY OBJECT > SECONDARY OBJECT INHERIl ^: RELATIONAL HIERARCHY Role: fingf [SUBJ> P. OBJ > S. OBJ] Syn: pers [ (1) 3 3 ] Sem: obv [ PROX+ ] 9.4.4 Sentential Obviation also operates across clauses. Within a sentence, the subject of a main clause can induce obviation on the subject of a subordinate clause. Consider the following sentence using the verb e-wabmat 'he sees him' where the main clause subject 211 is proximate and the subordinate clause subject is obviative: ^ (31) Bama zhe na mine e-wabmat [kwekseyen later EMPH EMPH again FCT-see . s . o\TA=3/3 ' C squirrel=OBV e-bmebtonet] . FCT-run . along\AI=0By=3C Later on, he (PROX) saw a squirrel (OBV) running along. [AS:2:2:021] To account for sentential obviation, we will need another hierarchy where main clause subjects outrank subordinate clause subjects, represented as follows: (32) SUBJECTS HffiRARCHY: SUB J > {SUB J} The Sentential Obviation Construction given in (33) will then inherit this hierarchy, and associate proximate with the highest ranked nominal, the main clause subject: (33) SENTENTIAL OBVIATION INHERIT: SUB.TECTS HIERARCHY Role: gf [SUBJ> {SUB J}] Syn: pers [3 3 ] Sem: obv [ PROX+ ] ' As a general rule, an independent verb that takes a complement inflects as if it had an inanimate object, or no object at all (so a TI or AI verb may be used). If the subject of the subordinate verb is animate, the independent verb may optionally inflect for an animate object. Some types of complement clauses do not allow this optionality, such as embedded content questions. However, some semantic classes of main clause verbs, such as perception verbs (as m this example), require the main clause verb to inflect for an animate object, if the complement has an animate subject. 212 A fully specified construct for the main clause verb in (3 1) is given on the following page, which shows how the Sentential Obviation construction can 'see' the subordinate clause subject. The external syntax and semantics are for the main clause verb, abbreviated as SEEING in the semantics. The three-part valence is my representation of subject-to-object copy, where the subject of the subordinate clause, 'SQUIRREL' is instantiated morphologically on the higher verb as primary object, and on the lower verb as subject. The subordinate clause subject is embedded in the valence of the subordinate clause verb RUNNING. 213 o < > PQ O u Pi < Pi w H O e^ PQ C/3 r" w w gas + + (U o r >^ "^1 Pi ^ a v^ r ft § 8 V a o a o ft, 5 a g 00 cd C a ■A S V. "^^ ^ C3 > o 2 ft ft t« bb o o a "" 1 m C4_i <X) 1 a a ia J r fin CM a o a- t/2 _ 1 s. f^ J vO S bO 2 ° a "" ^ \ J u 214 9.4.5 Sentence clusters The final syntactic domain for the operation of obviation is with sentence clusters, which as a less-common phenomenon, will only be briefly dealt with here. With sentence clusters (described in Section 8.4), a third person subject of one sentence can induce obviation of a third person subject in the following sentence, given a particularly close semantic relationship between the sentences. We capture this using a different hierarchy, given in (34). (34) SEQUENTIAL SUBJECTS HIERARCHY: SUBJ. > SUBJj This hierarchy will be inherited by the Sentence Cluster Obviation Construction, given in (35): (35) SENTENCE CLUSTER OBVIATION INHERIT: SEQ. SUB.IECTS HIERARCHY Role: gf [SUBJ, > SUBJj] Syn: pers [3 3 ] Sem: obv [ PROX+ ] 215 9.5 Discourse Obviation and the role of mental spaces We now turn to the use of obviation in discourse. When a narrator makes use of obviation at the discourse level, I will assume there is access to a default ranking of nominals relevant to the narrative (based on the narrator's global conception of his tale, goals in telling it, etc.). In the default ranking, the central character (the one the narrative is 'about') is ranked highest, and other characters are ranked lower depending on their importance in the narrative. A narrator may access other rankings at various points in the narrative, making another character a temporary proximate, which is known as a 'proximate shift'. However, the default ranking is the one predominantly used in the narrative, and the one to which a narrator will normally return after a proximate shift. The text I will be referring to in this section is 'Crane Boy' (given in Appendix C); in the previous chapter, I argued that the narrator made use of discourse obviation, which makes it suitable for analysis here. I will begin by constructing the default ranking of characters. The principle character is Crane Boy; he occurs early in the narrative, and is the central character in all subsequent episodes. The character he primarily interacts with is the Old Woman. Throughout the narrative, Crane Boy is generally maintained as a proximate, while the Old Woman is usually in the obviative. Other episodes that involve either Crane Boy or the Old Woman interacting with secondary characters have Crane Boy or the Old Woman as proximate, with the other characters as obviative. Based on the narrator's selection of proximates, we can rank the nominals in this narrative as follows: 216 (36) TOPIC HIERARCHY (DEFAULT, CRANE BOY NARRATIVE) Crane Boy > Old Woman > r Crane Boy's Parents, Bad Boy, Big Spoon, etc. > While this ranking is based on the proximate status of nominals in the narrative, it is also in accord with the overall topic structure; that is, the story is presented as being mainly about what happens to Crane Boy, and his experiences living with the Old Woman. Example (37) shows a mental spaces diagram that represents the act of narration in abbreviated form. The context of the narrative is represented by a space in the "Reality" Domain (Space R), and the narrative itself is represented by the space inside the Narrative Domain (Space N). Basic narration is 'external' narration, as opposed to 'internal' narration where the narrator adopts the viewpoint of one of the characters in the narrative. External narration, as shown in this diagram, takes place from the V-POINT of the narrator in the Reality Domain (see Chapter 7 for the representation of external narration in Mental Spaces theory). This V-POINT is associated with the default Topic Hierarchy, where Crane Boy outranks the Old Woman (The ranking is abbreviated here to include just Crane Boy and Old Woman.): 217 (37) DEFAULT TOPIC RANKING ASSOCIATED TO "REALITY" DOMAIN Space R: 'REALITY" DOMAIN BASE V-POINT TOPIC HIERARCHY: Crane Boy > Old Woman NARRATIVE DOMAIN Space N: FOCUS The constmction we will posit for Discourse Obviation (given in (38)) is similar to those already proposed for syntactic obviation. The primary difference is that it uses the Topic Hierarchy, which is a ranking of nominals based on their relative importance to the current discourse (the topic hierarchy is represented by ranking the nominal referents in 'Sem: ref in the abbreviated matrix). (38) DISCOURSE OBVIATION INHERIT: OBVIATION Sem: ref [ NOM. > NOMj ] Syn: pers \ 3 3 1 Sem: obv [ PROX+ ] 218 9.5.1 Inversion The Discourse Obviation construction is inherited, in turn, by the Direct and Inverse verbal constructions. To see how this works, we will examine a transitive verb. In a text such as Crane Boy, which shows evidence of proximate maintenance (and therefore a discourse topic hierarchy), we might expect to see a verb like the following (this example is constructed for ease of comparison with previous examples; there are plenty of comparable transitive verbs in the text). For our example, let us say that the notional subject is CRANE BOY and the notional object is OLD WOMAN (we will assume that the narrator 'chooses' which nominal referents will be associated with the notional subject and notional primary object). (39) e-wabmat e - wabEm -a -d FCT- see.s.o\TA -DIR -3C 'he [Crane Boy-PROX] saw her [Old Woman-OBV]' Let us also say that this example comes from a point in the text where there is external narration, that is, the narrator is not overtly representing the viewpoint of a character. The Topic Ranking in use is then the default ranking, which is available by the viewpoint of the external narrator, as shown in (37) above. The external semantics of the verbal construction references information about the mental spaces structure, such as the location of the BASE, V-POINT and FOCUS (shown in boldface in (40)). In this case, BASE and V-POINT are in "R" (the reality domain) and FOCUS is in "N" (the narrative domain). The location of V-POINT, in particular, provides access to the associated Topic Hierarchy. 219 (40) MENTAL SPACES ARE INDEXED INSIDE OF CONSTRUCTIONS INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT syn r cat V "N lex V r stem + ^ L word + J J val r. < syn role ^^ cat nom "^ per 3 ^CRANE BOY anim + prox - not gf subj fin gf subj 9 exp ~\ J frame SEEING part 1 part 2 spaces BASE R V-PODNT R FOCUS N u syn r cat nom "^ per 3 J OLD WOMAN anim + prox + V. role ( not gf pobj fin gf pobj 9 cont r\ ~\ J > V In this hypothetical example, the narrator has associated the notional subject with the final subject, motivating the use of the Direct Construction (the contribution of this construction is shown in boldface): 220 (41) DISCOURSE OBVIATION INHERITS DIRECT INHERIT: OBVIATION, DIRECT Sem: ref [ CRANE BOY > OLD WOMAN ] Role: not gf [ SUBJ > P.OBJ ] Role: fingf [ SUBJ > P.OBJ ] Syn: pers [ 3 3 ] Sem: obv [ PROX+ ] Now let us examine the case of an inverse verb. This time, the Old Woman is the notional subject and Crane Boy is the notional primary object: (42) e-wabmegot e - wabEm -EgO -d FCT- see.s.o\TA -INV -3C 'she [Old Woman-OBV] saw him [Crane Boy-PROX]' The space configuration remains the same, as for external narration: BASE and V-POINT are in R, and FOCUS is in N. This is shown in the external semantics of the verbal construct: 221 (43) MENTAL SPACES ARE INDEXED INSIDE OF CONSTRUCTIONS INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT syn r cat V lex r stem L word "N :] val r, < syn cat nom per 3 r CRANE BOY anim + prox + V- role f not gf pobj fm gf subj 9 exp V^ V y ~N y frame SEEING part 1 part 2 spaces BASE R V-POEVT R FOCUS N syn r cat nom "^ [ per 3 J r role OLD WOMAN anim + prox - not gf subj fin gf pobj 9 cont V. "^ y .^ "N y > V In this case, Discourse Obviation will still assign proximate status to Crane Boy as highest ranked nominal on the topicality scale and as the final subject. Old Woman, ranked lower on the topicality scale and the final primary object, will be obviative. Because there is a mismatch between notional and final relations as shown in the following matrix. Inverse applies: 222 (44) DISCOURSE OBVIATION INHERITS INVERSE INHERIT: OBVIATION, INVERSE Sem: ref [ CRANE BOY > OLD WOMAN ] Role: notgf [ P.OBJ < SUBJ ] Role: fin gf [ SUBJ > P.OBJ ] Syn: pers [3 3 ] Sem: obv [ PROX+ ] 9.5.2 Proximate shifts Narrators sometimes shift perspective to represent the viewpoint of a character. To do so, they access a V-POINT from within the narrative domain. Since V-POINT is associated with a Topic Hierarchy, accessing a different V-POINT can result in a proximate shift, where a secondary character is temporarily a proximate. To illustrate, in the Crane Boy narrative, there is a proximate shift when the Old Woman first hears Crane Boy crying, and approaches him (lines 15-18). During this episode, all references to the Old Woman are proximate, and the references to Crane Boy are obviative, which is expected if there is a 'rezeroing' of the center of deictic reference. The Topic Hierarchy linked to the Old Woman's viewpoint has Old Woman ranked highest, followed by Crane Boy (these are the only two characters in the episode): (45) TOPIC HIERARCHY (associated with Old Woman): Old woman > CraneBoy 223 In this case, we might expect that if Old Woman sees Crane Boy', Old Woman will be proximate, and Crane Boy obviative, reflecting the new Topic Hierarchy associated with the Old Woman. As such, a direct form would be used: (46) e-wabmat e - wabEm -a -d FCT- see.s.o\TA -DIR -3C 'she [Old Woman-PROX] saw him [Crane Boy-OBV]' This new topic hierarchy is indexed to a V-POINT inside the mental spaces network. The diagram in (46) shows a Character Domain inside of the Narrative Domain. This Character Domain represents the viewpoint (thoughts, construals, vantage point, etc.) of the Old Woman. The narrator, by representing the narrative as coming from the Old Woman's restricted point of view, makes use of 'internal' narration. This is represented in mental space terms by a V-POINT inside the Narrative Domain that is associated to the V-POINT of the Old Woman (represented by the arc in the diagram connecting the two "@" signs in each domain). This association link provides access to the Topic Hierarchy representing the Old Woman's viewpoint where Old Woman outranks Crane Boy: 224 (47) MENTAL SPACE REPRESENTATION OF A PROXIMATE SHIFT "REALITY" DOMAIN Space R: BASE TOPIC HIERARCHY: Old Woman> Crane Boy NARRATIl E DOMAIN CHARACTER DOMAIN (Old Woman) Space C The index to the new ranking is provided in the construct of the predicate, within the external semantics, as shown below (in boldface): 225 (48) NEW RANKING IS INDEXED INSIDE OF THE EXTERNAL SEMANTICS e-wabmat 'she [Old Woman-PROX] saw him [Crane Boy-OBV]' INHERIT: OBVIATION, DEFAULT OBVIATIVE ASSIGNMENT, DIRECT syn r V cat V lex ■> r stem + "^ L word + J J sem /^rame SEEING part 1 part 2 val < syn sem ^cat nom "^ per 3 prox + J OLD WOMAN g. an + J role r V. not gf subj fin gf subj exp -N J \^ spaces BASE R V-POEVT N FOCUS N syn role ^cat nom "^ per 3 prox - CRANE BOY g. an + not gf pobj fin gfpobj cont > In this case, proximate status will be assigned to Old Woman, as the final subject and highest ranking nominal on the new topic hierarchy. Crane Boy, lower on the hierarchy and the final primary object, will be obviative. The alignment of notional and final relations allows the Direct construction to apply: 226 (49) PROXIMATE SHIFT, DISCOURSE OBVIATION INHERITS DIRECT INHERIT: OBVIATION, DIRECT Sem: ref [ OLD WOMAN > CRANE BOY ] Role: not gf [ SUBJ > P.OBJ ] Role: fin gf [ SUBJ > P.OBJ ] Syn: pers [ 3 3 ] Sem: obv [ PROX+ ] 9.6 Discussion The sections above have presented an analysis of obviation in several domains: the phrase, the clause, within a sentence, sequential sentence clusters, and in discourse. I have argued that these uses of obviation are themselves constructions, which are related by shared inherance of the Obviation Construction. Besides sharing the inheritance of the Obviation Construction, these constructions are also similar to each other in the types of hierarchies they introduce. Although the hierarchies have been stated as determined by the morphological marking of obviation, there is reason to suspect a deeper similarity: An argument can be made for the overall saliency of higher ranked nominals, based on animacy (possession), agency (clausemate), syntactic embedding (sentential), semantic embedding (sentence clusters), and topicality (discourse). A likely motivation for the extension of Obviation in each case seems 227 therefore to be 1) non-coreferential third persons, and 2) a reasonable basis for estabUshing relative saliency among them. The Obviation Construction itself has a very broad function, that of linking proximate status with the highest ranking third person nominal on some unspecified hierarchy. Each construction that inherits Obviation adds information by contributing a specific hierarchy. A construction that makes use of this kind of inheritance relationship is known as an 'instance' construction (for a discussion, see Goldberg, 1995). The inheritance relationships for the Obviation instance constructions are shown in the diagram on the following page: 228 <+H o fl « o & -^ o o u fa ^ o ^■■B -*— » ote: ow Zi « fi 00 o l-H H U H on O u w u H oo >^ >! ^ JH o USTE ectsH i>^ 1 ION n Hier< in o •-F3 S3 TIAL Hierar CECL al Subj RSE erarcb SSESS ssessio S3 Z ^ SENTEN (Sequenti SCOU opic Hi 2e^ h-1 S S^ 1 1 ^ < 229 Returning to the question posed at the beginning of this chapter, by advocating a constructional approach, we assume that neither syntax nor discourse plays a more important role in the application of Obviation/>er se] that is, both Clausemate and Discourse Obviation represent polysemic extensions of the Obviation Construction. It is another question, however, which constructions a language has in its inventory, and the extent to which its speakers make use of them. In order to address this question, I propose the concept of 'constructional maintenance', where different languages, dialects (or even narrators!) may access a construction to varying degrees. With respect to Discourse Obviation, we might define the following degrees of maintenance (although I believe it to be essentially a cline): STRONG MAINTENANCE: the nominal highest in the topic rank will be the proximate within the discourse span. WEAK MAINTENANCE: attention to topic rank will be given in some contexts, generally more visible ones, but not others. NON-MAINTENANCE: the construction does not apply or is not available in the constructional inventory. Comparing languages then, we might represent the maintenance of Clausemate and Discourse Obviation as follows: 230 Clausemate Obviation Discourse Obviation Fox Weak to non-maintenance Strong maintenance Ottawa Strong maintenance Weak to non-maintenance Potawatomi Strong maintenance Weak maintenance A language like Fox has strong maintenance of discourse obviation, while having weak to non-maintenance of Clausemate Obviation. Ottawa is the reverse; it has strong maintenance of Clausemate Obviation, but weak to non-maintenance of Discourse Obviation. Potawatomi is somewhere in the middle of these extremes: it can be generally characterized as a syntax obviation language, with strong maintenance of Clausemate Obviation, however some narrators make limited use of discourse obviation (for instance in main clauses transitive verbs, but not with main clause intransitives), and so has weak maintenance of the Discourse Obviation Construction. 231 Bibliography Aissen, Judith. 1997. On the Syntax of Obviation. Language, 73.705-50. Anderson, Stephen R. 1992. A-Morphous Morphology.vol. 62: Cambridge Studies in Linguistics. Cambridge: Cambridge University Press. Bloomfield, Leonard. 1962. The Menomini Language. William Dwight Whitney Linguistic Series, ed. by ed. Charles F. Hockett: Yale University Press Dahlstrom, Amy. 1988. Plains Cree Morphosyntax, Department of Linguistics, University of California, Berkeley: Ph.D. — . 1 996. Narrative Structure in a Fox text, nikotwasik iskwahtem, paskihtepayih! Studies in Honour of H.C. Wolfart, ed. by John D.; Nichols and Arden Ogg. Winnepeg: Algonquian and Iroquoian Linguistics. Dunnigan, Tim, Patrick O'Malley and Linda Schwartz. 1978. Afunctional analysis of the Algonquian Obviative. Minnesota Papers in Linguistics and the Philosophy of Language, 5.7-22. Fauconnier, Gilles. 1985. Mental Spaces. New York: Cambridge University Press. Fleischman, S. 1990. Tense and Narrativity. Austin: University of Texas Press. Goddard, Ives. 1984. The obviative in Fox narrative discourse. Papers of the Fifteenth Algonquian Conference, ed. by William Cowan. Ottawa: Carleton University. — . 1990. Aspects of the Topic Structure of Fox Narratives: Proximate Shifts and the Use of Overt and Inflectional NPs. International Journal of American Linguistics, 56.317-40. Goldberg, A. 1995. Constructions: A Construction Grammar Approach to Argument Structure: Cognitive Theory of Language and Culture. Chicago: University of Chicago Press. Grafstein, Ann. 1981. Obviation in Ojibwa. Linguistique amerindienne II: etudes algonquiennes., ed. by Lynn Drapeau: Universite du Quebec a Montreal. Hockett, Charles. 1948a. Potawatomi I: Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948b. Potawatomi II: Derivation, Personal Prefixes, and Nouns. International Journal of American Linguistics, XIV.63-73. — . 1966. What Algonquian is really like. International Journal of American Linguistics, 32.59-73. Rhodes, Richard A. 1976. The Morphosyntax of the Central Ojibwa Verb, Linguistics, University of Michigan: Ph.D. — . 1985. Obviation: the Mark of Non-Coreference — . 1990. Obviation, Inversion, and Topic Rank in Ojibwa. Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society February 16- 19, 1990: Special Session on General Topics in American Indian Linguistics, 16S.101-15. — . 1992. The Syntax of Possessor Obviation in Ojibwe 232 — . 1994. Agency, Inversion, and Thematic Alignment in Ojibwe. Proceedings of the Twentieth Annual Meeting of the Berkeley Linguistics Society February 18-21, 1994: General Session: Dedicated to the Contributions of Charles J. Fillmore, ed. by Susanne; Dolbey Gahl, Andy; Johnson, Christopher, 431-46. Berkeley: Berkeley Linguistics Society. — . 2002. Obviation, Inversion and Topic Rank Revisited. Paper presented at Paper presented to the 34th Algonquian Conference, Kingston Ontario. Wolfart, H. Christoph. 1973. Plains Cree: A grammatical study. Transactions of the American Philosophical Society, 63. 233 10 Summary: Cross Domain Mappings 10.1 Cross-domain mappings According to Langacker, "semantic structures. . .are characterized relative to 'cognitive domains', where a domain can be any sort of conceptualization: a perceptual experience, a concept, a conceptual complex, an elaborate knowledge system, etc." (1991, p. 5). He gives the example of the predication 'knife' which requires at least a spatial domain (for its physical shape), one for the activity of cutting, and one for its membership in the set of silverware, and probably several others. The set of domains required for characterization of a predication he calls a "complex matrix." Within the matrix, a domain may be more or less central, based on the context in which it is used. Moreover, one domain may figure into other domains; in the 'knife' example, the spatial domain (which is probably more basic) a component of the other two. Similarly, a full description of the use of independents, conjuncts, the preverb e-, and obviation require reference to at least two grammatical domains.^ One domain is that of the sentence, and the other is that of discourse, which together constitute the complex matrix. And while we will need to talk about their use within different ^ I assume that linguistic knowledge constitutes a domain of experience. Within this wider domain, we have metalmguistic sub-domains or frames for sentence construction, the organization of discourse based on context, etc. 232 grammatical domains — that is, both in sentences and discourse — we will also need to characterize the relationships between their uses across these domains. 10.2 Cross Domain Mappings as Grammatical Blends Recent work in Mental Spaces theory has argued that blends are central to grammar (Fauconier and Turner, 1996; Fauconnier, 1997). Generally, these studies have focused on the idea that constructions are blends that combine an input Space 1 for the basic use of the construction with another Space 2 that provides a context for a plausible extension of the construction. When the blend is 'run', there is a mapping between counterparts in the two input domains, which are then projected into the blend. These common elements are projected into the blend. The form of the construction is also projected from input Space 1 allowing for the labeling of the construction with its new semantic extension. In this chapter, I will argue that the use of the independent, conjunct, preverb e- and obviation within syntax in everyday discourse, and their discourse uses in narrative are the result of grammatical blends. 10.2.1 Obviation The diagram in (1) represents the cross-space mapping for obviation, and shows how the marking of obviation in a particular domain might be extended to another domain that is perceived by speakers to be similar in semantic structure. 233 (1) OB VIATION BLEND Generic Space Input Spac^ 1 \ Discourse Structure ^ Input Space 2 Sentence Structure Blend The input spaces represent two obviation instance constructions (see Chapter 9). Input Space 1 contains a representation of Discourse Obviation, where jejakos 'Crane' outranks mdemoze 'Old Woman' on the Topic hierarchy (see Chapter 9, examples (39) - (41)). Input Space 2 contains a representation of Possessee Obviation where the possessor nene 'man' outranks the possessee wgwesen 'his son' (see Chapter 9 examples 234 (7) - (1 1)). In Input Space 1, the vertical arrow represents the topic hierarchy. This space has the most topical nominal jejakos 'Crane Boy' ranked higher than mdemoze 'Old Woman'. In Input Space 2, the vertical arrow represents the possession hierarchy, where the possessor outranks the possessee. This space shows a possessor, nene 'man' ranked higher than the possessee, wgwesen 'his son', (see Chapter 9 for a discussion). In the cross-domain mapping, the topic hierarchy maps onto the possession hierarchy. The highest ranking nominal in Input Space 1 maps onto the possessor in Input Space 2, and the lower ranked nominal in Input Space 1 maps onto the possessee in Input Space 2. The generic space represents the comparison of the input spaces, and contains a representation of the elements shared by the input spaces. In this case, the generic space contains a hierarchy, non-coreferential third persons "A" and "B" that are ranked relative to the hierarchy, and requires the grammatical marking of the lower ranked nominal. This, in fact, is a good representation of the Obviation Construction. Once the mapping between the elements of the input spaces is established, the blend can be 'run'. The blend contains the hierarchy of ranked possessor and possess from the Input Space 2, and takes the grammatical marking of the lower ranked nominal from Input Space 1 . The result is grammatical marking of obviation in a new syntactic domain. 10.2.2 Main Clause Conjunct Verbs in Narrative The next case I will consider is the use of main clause conjuncts in narrative discourse. I have argued that the use of the conjunct in the main clauses of narrative 235 foreground sentences (that is, the NC) represents the embedding, or subordination, of narrative within a larger non-narrative discourse (see Chapter 6). I argue below that this is also accomplished with a blend. The set up of the blend is much the same as for obviation, with both a syntactic and discourse input space. In the diagram in (2), Input Space 1 (Sentence Structure) contains a representation of a complex sentence with a subordinate clause, the subordinate clause containing a conjunct verb, indicated with a "C" (the argument for this type of representation is given in Chapter 4). Input Space 2 (Discourse Structure) contains a representation of narrative discourse embedded inside of a larger non-narrative discourse. The line between the two spaces represents a division of information into the "Reality" Domain (everyday discourse) and the Narrative Domain (narrative discourse). (The argument for this representation is given in Chapter 7.) 236 (2) NARRATIVE CONSTRUCTION BLEND Input Space 1 ^\v^ Sentence Structure \ Generic Space "Reality" Domain Narrative Domain Input Space 2 /' Discourse Structure Blend The cross-space mappings are as follows: The main clause of Sentence Structure maps onto the "Reality" Domain network of Discourse Structure, the subordinate clause of Sentence Structure maps onto the Narrative Domain network of Discourse Structure. The Generic Space contains a representation of a complex structure with a parent space and a subordinate space. The parent space maps onto the main clause in Sentence 237 Structure and the "Reality" Domain network of Discourse Structure, and the subordinate space maps onto the subordinate clause in Sentence Structure and the Narrative Domain network in Discourse Structure. The blend functions to map the subordinate clause from Sentence Structure onto the Narrative Domain network from Discourse Structure, and crucially provides the label — the conjunct — which is then applied to the Narrative Domain network. The way the conjunct specifically represents the Narrative Domain network is accomplished through a series of metonymies, as follows: (3) Main clause Instance of Narrative Narrative conjunct -^ a sentence -^ foreground ^ discourse in the NC information In each mapping, the smaller grammatical unit stands for the larger unit that includes it: the main clause stands for a sentence as a whole, so a main clause conjunct can stand for a sentence in the NC pattern. The NC pattern represents narrative foreground information, and this in turn represents narrative discourse. (The main clause conjunct alone does not trigger this mapping, since there are other uses of main clause conjuncts, as described in Chapter 4. I presume there are many contextual cues along with the use of main clause conjuncts that indicate a narrative discourse). 238 10.2.3 Main Clause Independent Verbs in Narrative The use of main clause independent verbs in narrative (that is the use of the CC) can also be represented as a blend, in much the same way as with the use of main clause conjuncts in narrative, as shown in (4): (4) CONVERSATIONAL CONSTRUCTION BLEND Input Space 1 \ "^-^^ Sentence Structure\ Narrative Domain Input Space 2 Discourse Structure Blend 239 In this case, the critical cross-space mapping is of the main clause independent verb (represented by "I" in the Sentence Structure input space) onto the "Reality" Domain network of Discourse Structure. In the blend, the use of the independent "label" gets extended to the "Reality" Domain. As with the conjunct, there is a series of metonymies: (5) Main clause Instance of Everyday independent -^ a sentence -^ discourse in the CC A main clause independent stands for a sentence in the CC pattern, which in turn is representative of everyday discourse. This indexicality of the independent for the "Reality" Domain is not as apparent as the indexicality of the conjunct for the Narrative Domain, largely because this function is 'hidden in plain view'. That is, it takes the contrast of narrative sentences with main clause conjuncts to show this functionality of the conjunct. Everyday discourse does not, in and of itself, show the indexicality of the independent for a non-embedded domain. Clues to this use are, however, provided by the use of main clause independent verbs in narrative (instances of the CC), which I have argued index everyday discourse (and thus the "Reality" Domain) in various ways. The types of information that the CC can represent due to this indexicality include background and focalized information — see Chapter 6 for a discussion). 240 10.2.4 The e- preverb In Chapter 4, we argued that the function of the e- preverb within the sentence is as a kind of factive, indicating strong speaker confidence in the factuality of the proposition expressed in a subordinate clause. In Chapter 6, we argued for its role in narrative discourse as a kind of evidential, marking the strong epistemic stance conventionally taken by a speaker in the telling of a traditional narrative. We are now able to demonstrate that the use of the e- preverb on conjunct verbs in the main clauses of narrative foreground sentences is another instance of a blend that takes Sentence Structure and Discourse Structure as input spaces. Since the e- preverb accompanies main conjuncts in narrative foreground sentences, it makes sense to use the basic blend structure given in (2) for the use of main clause conjuncts in narrative. The blend for the e- preverb is shown in (6) below: 241 (6) BLEND FOR THE PREVERB E- Input Space 1 "■\ Sentence Structure Generic Space 'Reality" Domain Narrative Domain Input Space 2 /' Discourse Structure Blend The input spaces are again Sentence Structure and Discourse Structure. The subordinate clause space in Sentence Structure maps onto the subordinate Narrative Domain in Discourse Structure. In the blend, the label "e-" is applied to the Narrative Domain, to which it contributes its semantics as a marker of factivity. Its association to 242 main clause conjuncts as representative of the narrative domain is again accomplished through the series of metonymies as given in (7), repeated and slightly modified below: (7) Main clause Instance of Narrative Narrative e-conjunct -^ a sentence -^ foreground ^ discourse in the NC information 10.2.5 Directionality of mapping In the discussion above, I have represented particular mapping as being projected from one domain onto another. While I assume that there is a directionality to the mapping, I am not here making a claim about the particular directionality of each blend. The directionality I have posited for the purposes of exposition are merely those that seem to be plausible directions of grammaticalization in each case. That is, it seems plausible that the 'basic' uses of independents, conjuncts and the preverb e- are what we find in everyday discourse, and their narrative uses are derived from this. However, knowing that conjuncts are older verb forms and narrative discourse tends to be conservative, there are likely good arguments for the opposite directionality.^ Obviation, on the other hand, more likely arose as a discourse mechanism, and seems to be grammaticalized in syntax (in fact, I have argued that Potawatomi shows this in progress). My point is the mapping could go either way without undermining the existence of the blend. While I find the question of directionality an interesting one, I See Goddard (1967) for a discussion of the development of independent verbs as nominalizations. 243 am mainly interested in establishing the existance of a mapping between the domains of discourse and syntax. 10.3 Conclusion The descriptive problem posed at the beginning of this discussion was the behavior of several grammatical elements in Potawatomi whose distributions vary depending on the discourse context. Standard theories of syntax are bound to fail at an explanation, because they cannot "see" the discourse. Without reference to discourse, how can one reconcile the fact that in Potawatomi, both independent and conjunct verbs are used to mark main clauses? Or that the preverb e- has a restricted use in conversation to certain types of subordinate clauses, but proliferates to nearly every finite verb in narrative? I have argued that a cognitive linguistic framework provides the means of describing such constructions whose distribution is dependent on discourse context. Using the theory of Mental Spaces, I have argued that these different distributions represent constructional polysemy, where a single grammatical form is mapped onto multiple functions. Because discourse structures are seen as part of a continuum of form- meaning pairings that include syntactic structures, it makes sense that functions of constructions might be predicated in these different domains. In this chapter, I have argued for the existence of several mental space blends in Potawatomi that take as their inputs constructions in syntax and constructions in discourse. Existing contexts for the use of a construction are compared to possible new contexts, and this comparison generates cross-space mappings. If there are enough 244 similarities, and the motivation is strong enough, the new context may be adopted, the blend run, and the marking (form) of the construction can be extended to the new, semantically related function. While I have argued that this blend structure is productive in Potawatomi, it seems likely to be productive in many, if not most languages, given the assumption that in all languages, syntactic structures and discourse structures are the same basic kinds of entities. The goals of this dissertation were to describe several areas of Potawatomi morphosyntax that have not been given much attention in the literature, and at the same time to argue for a theory of grammar that allows an examination of relationships across traditional domains of grammatical description. I have argued that the use of independents, conjuncts, the preverb e- and obviation have functions across grammatical domains, and that a full grammatical description requires not only addressing their use in each domain, but the relationship between their functions across these domains. Since each discourse genre comes with a set of requirements about grammatical form, it makes sense to describe grammatical form with reference to those genres. And, only after we can talk about this relationship can we address the possibility of a systematicity to the various uses of these constructions. 245 Fauconier, Gilles and Mark Turner. 1996. Blending as a central process of grammar. Conceptual Structure, Discourse and Language, ed. by Adele Goldberg. Stanford: CSLI. Fauconnier, Gilles. 1997. Mappings in Thought and Language. Cambridge: Cambridge University Press. Goddard, Ives. 1967. The Algonquian Independent Indicative. National Museum of Canada Contributions to Anthropology, Linguistics 1.66-106. Langacker, R. 1991. Concept, Image, and Symbol: The Cognitive Basis of Grammar. vol. 1: Cognitive Linguistics Research. Berlin: Mouton de Gruyter. 246 Bibliography AISSEN, JUDITH. 1997. On the Syntax of Obviation. Language, 73.705-50. ANDERSON, STEPHEN R. 1992. A-Morphous Morphology .vol. 62: Cambridge Studies in Linguistics. Cambridge: Cambridge University Press. BLOOMFIELD, LEONARD. 1927. Notes on the Fox Language. International Journal of American Linguistics, 4.181-219. — . 1946. Algonquian. Linguistic Structures of Native America, ed. by H. Hoijer, 85-129. New York. — . 1958. Eastern Ojibwa: Grammatical Sketch, Texts and Word List. Ann Arbor: University of Michigan Press. — . 1962. The Menomini Language. William Dwight Whitney Linguistic Series, ed. by ed. Charles F. Hockett: Yale University Press COSTA, DAVID. 1996. Reconstructing Initial Change in Algonquian. Anthropological Linguistics, 38.39-72. CUOQ, JEAN- ANDRE. 1866. Etudes philologiques sur quelques langues sauvages de I'Amerique. Montreal: Dawson Brothers. CUTRER, MICHELLE. 1994. Time and Tense in Narratives and Everyday Language, University of California: Ph.D. dissertation. DAHLSTROM, AMY. 1988. Plains Cree Morphosyntax, Department of Linguistics, University of California, Berkeley: Ph.D. 246 — . 1996. Narrative Structure in a Fox text, nikotwasik iskwahtem, paskihtepayih! Studies in Honour of H.C. Wolfart, ed. by John D.; Nichols and Arden Ogg. Winnepeg: Algonquian and Iroquoian Linguistics. DANCYGIER, BARBARA; and EVE SWEETSER. 1996. Conditionals, Distancing, and Alternative Spaces. Conceptual Structure, Discourse, and Language, ed. by Adele Goldberg, 83-99. Stanford: CSLI Publications. DRYER, MATTHEW. 1992. A Comparison of the Obviation Systems of Kutenai and Algonquian. Papers of the Twenty-Third Algonquian Conference., ed. by William Cowan, 119-63. Ottawa: Carleton University. DUNNIGAN, TIM, PATRICK O'MALLEY and LINDA SCHWARTZ. 1978. A functional analysis of the Algonquian Obviative. Minnesota Papers in Linguistics and the Philosophy of Language, 5.7-22. FAUCONIER, GILLES and MARK TURNER. 1996. Blending as a central process of grammar. Conceptual Structure, Discourse and Language, ed. by Adele Goldberg. Stanford: CSLI. FAUCONNIER, GILLES. 1984. Espaces mentaux. Paris: Minuit. — . 1985. Mental Spaces. New York: Cambridge University Press. — . 1997. Mappings in Thought and Language. Cambridge: Cambridge University Press. FILLMORE, CHARLES and PAUL KAY. 1993. Construction Grammar, vol. Unpublished ms. Berkeley, CA — . 1999. Grammatical constructions and linguistic generalizations: The What's X doing Y? Construction. Language, 75.1-33. FLEISCHMAN, SUZANNE. 1990. Tense and Narrativity. Austin: University of Texas Press. 247 FRANTZ, DONALD. 1966. Person indexing in Blackfoot. International Journal of American Linguistics, 32.50-58. GENETTE, GERARD. 1980. Narrative Discourse an Essay in Method. Ithaca, New York: Cornell University Press. GODDARD, IVES. 1967. The Algonquian Independent Indicative. National Museum of Canada Contributions to Anthropology, Linguistics 1.66-106. — . 1974. Remarks on the Algonquian Independent Indicative. International Journal of American Linguistics, 40.317-27. — . 1984. The obviative in Fox narrative discourse. Papers of the Fifteenth Algonquian Conference, ed. by William Cowan. Ottawa: Carleton University. — . 1990. Aspects of the Topic Structure of Fox Narratives: Proximate Shifts and the Use of Overt and Inflectional NPs. International Journal of American Linguistics, 56.317-40. GOLDBERG, ADELE. 1995. Constructions: A Construction Grammar Approach to Argument Structure: Cognitive Theory of Language and Culture. Chicago: University of Chicago Press. GRAFSTEIN, ANN. 1981. Obviation in Ojibwa. Linguistique amerindienne II: etudes algonquiennes., ed. by Lynn Drapeau: Universite du Quebec a Montreal. GRIMES, JOSEPH E. 1975. The Thread of Discourse, vol. 207: Janua Linguarum Series Minor. The Hague: Mouton. HOCKETT, CHARLES. 1939. The Potawatomi Language, Anthropology, Yale University: Ph.D. dissertation. — . 1939. Potawatomi Syntax. Language, 15.235-48. 248 — . 1948. Potawatomi I: Phonemics, Morphophonemics, and Morphological Survey. International Journal of American Linguistics, XIV. 1-10. — . 1948. Potawatomi II: Derivation, Personal Prefixes, and Nouns. International Journal of American Linguistics, XIV.63-73. — . 1948. Potawatomi III: The Verb Complex. International Journal of American Linguistics, XIV. 139-49. — . 1948. Potawatomi IV: Particles and Sample Texts. International Journal of American Linguistics, XIV.213-25. — . 1966. What Algonquian is really like. International Journal of American Linguistics, 32.59-73. HOPPER, PAUL. 1979. Aspect and Foregrounding in Discourse. Discourse and Syntax, ed. by Talmy Givon, 213-41. New York: Academic Press. — . 1979. Some observations on the typology of focus and aspect in narrative language. Studies in Language, 3.37-64. HOPPER, PAUL J. and SANDRA A. THOMPSON. 1980. Transitivity in Grammar and Discourse. Language, 56.251-99. JACKENDOFF, RAY. 1975. On belief contexts. Linguistic Inquiry, 6. JACOB SEN, WILLIAM. 1967. Switch-Reference in Hokan-Coahuiltecan. Studies in Southwestern Ethnolinguistics, ed. by Dell H. Hymes, 238-63. The Hague. JAMES, DEBORAH. 1983. Simple versus Conjunct Verbs in Moose Cree: Some Whys and Wherefores. Actes du Quatorzieme Congres des Algonquinistes, ed. by W. Cowan, 345-61. Ottawa: Carleton University. JESPERSON, OTTO. 1924. The Philosophy of Grammar. Chicago: University of Chicago Press. 249 LABOV, WILLLIAM and JOSHUA WALETZKY. 1967. Narrative analysis. Essays on the Verbal and Visual Arts, ed. by J. Helm, 12-44. Seattle: Unversity of Washington Press. LABOV, WILLIAM. 1972. Language in the Inner City: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. LAKOFF, ROBIN. 1968. Abstract Syntax and Latin Complementation. Cambridge: The MIT Press. LAKOFF, GEORGE. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. Chicago: University of Chicago Press. LANGACKER, RONALD. 1991. Concept, Image, and Symbol: The Cognitive Basis of Grammar .vol. 1: Cognitive Linguistics Research. Berlin: Mouton de Gruyter. LEVINSON, STEPHEN C. 1983. Pragmatics: Cambridge Textbooks in Linguistics. Cambridge: Cambridge University Press. LIDDELL, SCOTT. 1995. Real, Surrogate, and Token Space: Grammatical Consequences in ASL. Language, Gesture, and Space, ed. by J. Reilly, 19-41. New Jersey: Laurence Erlbaum Assoc. LONGACRE, ROBERT. 1976. 'Mystery' particles and affixes. Proceedings of the Twelfth Regional Meeting of the Chicago Linguistic Society. 468-75. — . 1985. Sentences as combinations of clauses. Language typology and syntactic description. Vol. II: Complex Constructions, ed. by Timothy Shopen, 235-86. Cambridge: Cambridge University Press. MICHELSON, TRUMAN. 1921. The Owl Sacred Pack of the Fox Indians.vol. Bulletin no. 72: Bureau of American Ethnology. Washington, D.C.: Smithsonian Institution. 250 — . 1925. Accompanying papers.vol. 40: Bureau of Emerican Ethnology Annual Report. Washington, D.C.: Smithsonian Institution. MITHUN, MARIANNE. 1990. The Languages of Native North America. Cambridge: Cambridge University Press. MUSHIN, ILANA. 1998. Viewpoint Shifts in Narrative. Discourse and Cognition, ed. by Jean- Pierre Koenig, 323-36. Stanford: CSLI Publications. NOONAN, MICHAEL. 1985. Complementation. Language typology and syntactic description. Vol. II: Complex Constructions., ed. by Timothy Shopen. Cambridge: Cambridge University Press. NUNBERG, GEOFFREY. 1978. The pragmatics of reference. Bloomington, Ind.: Indiana University Linguistics Club. — . 1979. The non-uniqueness of semantic solutions: Polysemy. Linguistics and Philosophy, 3. OCHS, ELINOR. 1979. Planned and Unplanned Discourse. Discourse and Syntax: Academic Press, Inc. PLATO. 1968. The Republic. New York: Basic Books. RHODES, RICHARD A. 1976. The Morphosyntax of the Central Ojibwa Verb, Linguistics, University of Michigan: Ph.D. — . 1985. Eastern Ojibwa-Chippewa-Ottawa Dictionary. Berlin: Walter de Gruyter & Co. — . 1985. Obviation: the Mark of Non-Coreference — . 1990. Obviation, Inversion, and Topic Rank in Ojibwa. Proceedings of the Sixteenth Annual Meeting of the Berkeley Linguistics Society February 16-19, 1990: Special Session on General Topics in American Indian Linguistics, 16S. 101-15. — . 1992. Ottawa Word Order. Paper presented at SSILA Summer Meeting. 251 — . 1992. The Syntax of Possessor Obviation in Ojibwe — . 1993. The Possessor Constraint — . 1994. Agency, Inversion, and Thematic Alignment in Ojibwe. Proceedings of the Twentieth Annual Meeting of the Berkeley Linguistics Society February 18-21, 1994: General Session: Dedicated to the Contributions of Charles J. Fillmore, ed. by Susanne; Dolbey Gahl, Andy; Johnson, Christopher, 431-46. Berkeley: Berkeley Linguistics Society. RHODES, RICHARD A. 2002. Obviation, Inversion and Topic Rank Revisited. Paper presented at Paper presented to the 34th Algonquian Conference, Kingston Ontario. SANDERS, JOSE and GISELA REDEKER. 1996. Speech and Thought in Narrative Discourse. Spaces, Worlds, and Grammar, ed. by G. Fauconnier and E. Sweetser. Chicago: The University of Chicago Press. THOMPSON, SANDRA; and ROBERT LONGACRE. 1985. Adverbial clauses: Language typology and syntactic description. Vol. II Complex constructions. Cambridge: Cambridge University Press. THUNDER, JIM SR. and KIM WENSAUT. 1998. Bodewadmimwen Nswe Mbok (Bodewadmi Language Book Three): Jim Thunder and Kim Wensaut, publishers. UHLENBECK, CHRISTIANUS CORNELIUS. 1909. Grammatische onderscheidigen in het Algonkinsch. Amsterdam: Akademie van Wetenschappen. VALENTINE, J. RANDOLPH. 2001. Nishnaabemwin Reference Grammar. Toronto: University of Toronto Press. WOLF ART, H. CHRISTOPH. 1973. Plains Cree: A grammatical study. Transactions of the American Philosophical Society, 63. 252 Appendix A Grammatical Codes Used in Morpheme Glosses AI intransitive verb stem that takes an animate subject AN animate AUG augment suffix (optional with some II stems) C verb stem inflected in the conjunct paradigm CH initial change DIM diminutive suffix DIR direct theme suffix DUB dubitative suffix EMPH emphatic particle FCT factive prefix PUT future tense prefix I verb stem inflected in the independent paradigm II intransitive verb stem that takes an inanimate subject IMP verb stem inflected in the imperative INAN inanimate INV inverse theme suffix LOC locative suffix MOD modal prefix NEG negative suffix OBJ object suffix (on some TI stems) P verb stem inflected in the participle PASS passive suffix PL plural suffix PRET preterite suffix PST past tense prefix TA transitive verb stem that takes an animate object TI transitive verb stem that takes an inanimate object Person inflection ; 1 first person 2 second person 3 third person 3' third person obviative 15 first person plural, exclusive 12 first person plural, inclusive 25 second person plural 35 third person plural inanimate 05 inanimate plural X indefinite 253 Appendix B. Glossed Examples from Chapters 6 and 7 1 Examples from Chapter 6 (1) (JS.4.1) 1 I me se ngodek neshnabek iw mE sE nEgOd -Eg EnEshEnabe -g that.INAN EMPH EMPH one -LOG person -PL e-wdodanwat i je weye e- wEdodanE -wad iw jE weye FCT- have . a . village\AI -35. C and someone e-nshonajtagwat wgetgansewan e - nEshwEna jEtEw -Egwat wE- gEtEganEs -Ewan FCT- destroy. s.t. on. s.o. \TA -3'/35.C 3- garden -35 mine mbish wed'emwat. minE nEbish CC.OdE'EmEw -ad and water CC . get . it . from. there . for . s . o . \TA -35. P Once there was a village, and someone was destroying their gardens and wells. 2 Iw je nish wshkalsewsen e-gi-nokanawat iw jE nizh OshkabewEs -En e- gi- nokaN -awad and two helper -OBV FCT- PST- have . s . o . do . s . t . \TA -35/3 '.C e-wi-kewabmawat wegwendek o e- wi- EkEwabEm -awad wegwen -EdEg ow FCT- FUT- watch. out. for. s.o. \TA -35/3 '.C whomever -DUB that. AN ezhcheget . CC.EzhEchEge -d CC . do . things . a . certain . way\AI -3P So they had two scouts watch out for whomever might be doing that. 3 I je bama zhe na geten e-byanet weye. iw jE bama zhE na geten e- bya/e -nEd weye and later EMPH EMPH sure FCT- come\AI -3'.C someone Later, sure enough, someone came along. 4 E-wabmawat kojesen e-bshkobnanet; e- wabEm -awad kojes -En e- bEshkobEn -anEd FCT- see.s.o\TA -35/3 '.C bean -OBV FCT- pull . out . s . o . \TA -3/3 '.C 254 jak zhena e-zhechgenet . jag EzhEna e- EzhEchEge -nEd all EMPH FCT- do . things . a . certain . way\AI -3'.C They saw him pulling out beans and doing all kinds of things. 5 Wabozoyen je ni . wabozo#y -En jE niw rabbit -OBV but that.OBV It was Rabbit. (2) (JS.1.1) 1 Ngodek walsgonoshkwe e-gche-mwet jik-zibe. nEgOd -Eg wabEgonoshkwe e- gEchE- mEw -Ed jig- zibE one -LOG rat FCT- really- cry\AI -3.C next. to- river Once a rat was crying by the edge of a river. (3) (JS.4.4) 1 ode yadsokan espen e-bmebtot. odE yadsokan esEbEn e- bEmEbEto -d this story raccoon FCT- run.along\AI -3.C This story is about the Raccoon running along. (4) (JS.4.5) 1 I me se ngodek neshnabek iw mE sE nEgOd -Eg EnEshEnabe -g that.INAN EMPH EMPH one -LOC person -PL e-wdodanwat . e- wEdodanE -wad FCT- have.a.village\AI -35. C Once there was a village. (6) (JS.4.2) 50 E-bwamshe-nyewgongek e- bwamEshE- nyewgonEg -Eg FCT- before- be . f our . days\II -O.C e-byawat giw neyap i je o nene e- bya/e -wad giw neyab iw jE ow EnEnE#w FCT- come\AI -35C those. AN back and that. AN man 255 e-nat niw osen, e- En -ad niw #os -En FCT- say. to. s.o\TA -3/3'. C that . OBV father -OBV "Nnedwendan debendemak. " nE- nEdwend -a -n CH.dEbEnd -Emag want.s.t\TI -OBJ -OBJ own.s.t.\TI -15/0. C e-nat. e- En -ad FCT- say. to. s.o\TA -3/3'. C Before the four days ended, the couple came back, and the man said to his father, "I want our belongings. " (7) (MD 102694) 46 Iw je i ga-nakwneget iw jE iw CH.gi- nakOnEge -d and that.INAN CH.PST- plan . things\AI -3.P e-wi-debmat pi bwamshe e- wi- debEm -ad Opi bwamEshE FCT- FUT- grab.s.o\TA -3/3 '.C when before gwabtonet . '^ gwabEto -nEd run . out . of . water\AI -3 ' . C The one that planned it would grab him before he reached the shore. (8) (MD 102694) 28 A, babwichget jigbyek. a babwichEge -Ed jigbyeg ah wait . for . things\AI -3.C by. the. water Ah, he waited there by the shore. (9) (MD 102694) 35 A, gkanabmat o wabozo. a gEkanabEm -ad ow wabozo#y ah look. over. at. s.o\TA -3/3'. C that. AN rabbit Ah, the rabbit looked across at him. ^ This is a conjunct, not a participle. The participle would be egwabtot. 256 (10) (JS.4.1) 44 Iw je e-bme-byat niw iw jE e- bEmE- bya/e -d niw and FCT- in . the . process . of- come\AI -3.C that.OBV beshkmwen e-nat "Nseze! bEshkEmwe -n e- En -ad nE- #sEze lion -OBV FCT- say . to . s . o\TA -3/3'. C 1- older . brother Gyetnam nzeges . gyetnam nE- zegEzE for. sure 1- be . scared\AI . I When he [Rabbit] came across the lion he said to him, "Brother, I'm very scared. 45 Nwebi ' we . nE- webi ' Ewe 1- run . away . f rom. people\AI . I I'm running away from someone. 46 Weye zhode nshiwnagze anwe ge gin weye zh odE nEshiwnagOzE anwE ge gin someone here look . hostile\AI . 3 . I all. right also 2 . EMPH gneshiwnagwes nesh je win nwech. gE- nEshiwnagOzE nEzh jE win nwEj 2- look. hostile\AI . 3 . I contrarily but 3 . EMPH more There is someone here pretty scary; and you're scary, but he's even worse. 47 Ibe gge-zhyamen; geten nshiwnagze." ibE gE- gE- Ezhya/e -mEn geten nEshiwnagOzE there 2- FUT- go.there\AI -15.1 sure look . hostile\AI . 3 . I Let's go over there; he sure is scary. " 48 Beshkmwe e-kedot, "Gzhyamen, bEshkEmwe e- EkEdO -d gE- Ezhya/e -mEn lion FCT- say\AI -3 . C 2- go.there\AI -15.1 gge-we-wabmamen. " gE- gE- wE- wabEm -a -mEn 2- FUT- go. and- see.s.o\TA -DIR -12.1 Lion said, 'Lets go and take a look at him. " 257 (11) (HOPTl) 2 Neshnabe je o weni'get esben EnEshEnabe jE ow CC.wEni'Ege -d esEbEn person but that. AN trap -3.P raccoon gi -yawe . gi- YawE PST- be. a. certain. thing\AI . 3 . 1 When thelndian went trapping, the raccoon went along. (12) (HOPTl) 11 Gi je yaygenwik je giw jE YayEgenO -g jE those. AN but be . the . same . size\AI -35.1 but giw, jo je mamda giw jo jE mamda those. AN not but possible e-wi-wepodwat, e- wi- wepod -Ewad FCT- FUT- hit.s.t.\TI -3/3 '.C e-bwa-gkenmat ni e- bwa- gEkenEm -ad niw FCT- NEG- know.s.o.\TA -3/3'. C that . OBV wde-esbenmen . wEdE- esEbEn -Em -En 3- raccoon -POSS -OBV They were just the same size, these two, you see; so it was impossible for him [the man] to hit him [the other coon]; he couldn't tell which one was his own. 12 Pene je niw wde-esbenmen pEne jE niw wEdE- e sEbEn -Em -En always but that. OBV 3- raccoon -POSS -OBV nam-yegwan gi-wjeshnon. nam-yEgwan gi- OjEshEn -on under . something PST- lie . underneath\AI -3 '.I His own coon was always underneath. 258 (13) (JS.4.4) 1 Ode yadsokan espen e-bmebtot. odE yadsokan esEbEn e- bEmEbEto -d this story raccoon FCT- run.along\AI -3.C This story is about the Raccoon running along. 2 E-ye-bmebtot o espen e- ye- bEmEbEto -Ed ow esEpEn FCT- while^- run.along\AI -3.C that. AN raccoon wgi-wabman amon e-gojnenet. w- gi- wabEm -a -n amo -n e- gojEn -EnEd 3- PST- see.s.o\TA -DIR -OBV bee -PL FCT- hang\AI -3'.C While Raccoon was running along, he saw bees (a hive) hanging (from a tree). 3 Ga-zhewebzet je CC.gi- EzhEwebEzE -Ed jE PST- have . happen . to . one . a . certain . way\AI -3.C but gi-gmegmode gokosh wzheyen gi- gEmEgEmode gokosh w- #zh Ey -En PST- RED. steal. s. t. \AIO. 3. I pig 3- skin -OBV ngo ji . ngOj i somewhere He would go about stealing pork rind somewhere. (14) (JS.4.5) 1 I me se ngodek neshnabek iw mE sE nEgOd -Eg EnEshEnabe -g that.INAN EMPH EMPH one -LOC person -PL e-wdodanwat . e- wEdodanE -wad FCT- have.a.village\AI -35. C Once there was a village. ' This preverb, which is not attested elsewhere in the corpus, may be some kind of initial change. 259 Gi-dbedbowek; gego zhena gi- dEbEdEbEwe -g gego EzhEna PST- DUP.have. a. council\AI -3.C something EMPH gi-ya jdanawat . gi- YajEd -a -n -awad PST- tell.of . s. t\TI -OBJ -OBJ -35/3'. C They were having a council; talking about something. I je ibe mbesek nawesh [gagita] odan iw jE ibE nEbEs -Eg nawezh odan and there lake -LOG in . the .middle . of . an . area town gi -yawen ibe . gi- yawEn ibE PST- be . in . a . certain . place\II . . I there And there was a town in the middle of a lake. 4 I je ye i ga-wje-dbedbowewat . iw je ye iw CC.gi- wEjE- dEbEdEbEwe -wad that ' s. the. one. (INAN) CO. PST- where- DUP . have . a . council\AI -35. P That's where they would go for their council. 5 I je ngot nene neshzhena gi-wijewe iw jE nEgOd EnEnE#w nEshzhEna gi- wijewe and one man for. no . reason PST- go . along.with.people\AI . 3 . I neko . nEko used. to So there was one man who used to go along for no particular reason. 6 Jo zhena win gego jo EzhEna win gego not EMPH 3. EMPH something gi-zhe-dbowesi neshzhena gi- EzhE- dEbEwe -si nEshzhEna PST- in . a . certain . way- have . a . council\AI -NEG.I f or . no . reason e-zhyat . e- Ezhya/e -d FCT- go.there\AI -3.C He did not go for the council; he went for no particular reason. 260 7 Ga-wje-zhyat je e-wi-mnekwet . CC.gi- wEjE- Ezhya/e -d jE e- wi- mEnEkwe -d CC.PST- why- go.there\AI -3.P but FCT- FUT- drink\AI -3 . C The reason he went was to drink. 8 Ngodek e-dokit bama zhena jo weye; jayek nEgOd -Eg e- doki -d bama EzhEna jo weye jayeg one -LOG FCT- wake.up\AI -3 . C later EMPH not someone all gi -ma j iwagben . gi- maji -wag -EbEn PST- leave\AI -3 -PRET.I Once this man woke up and nobody was there; everyone must have left. 9 E-gingenayek nsheke. e- ginEgEn -ayE -g nEshEke FCT- leave. s . o. alone\TA -PASS -3 . C alone He was left all alone. 10 Ngodek jigbyek e-gi-we- jajibdebet nEgOd -Eg jigbyeg e- gi- wE- DUP.jibEdEbE -d one -LOC by. the. water FCT- PST- go . and- sit\AIO -3.C gdewanen e-giwadzet i j® o mtek gEdEwan -En e- giwadEzE -d iw jE ow mEtEg log -OBV FCT- be.lonely\AI -3 . C and that. AN tree e-gi-ggenonat . e- gi- gEgEnon -ad FCT- PST- talk. to. s.o. \TA -3/3'. C One time he went by the lake and sat by a log, feeling lonely, and the tree talked to him. (15) (JS.4.1) 88 I je iw bodwagen megwa shkode iw jE iw bodwagEn megwa Eshkode and that.INAN fireplace still fire gi-temget . gi- te -mEgEd PST- be. in. a. certain. place -AUG. O.I So there was still a fire in the fireplace. 261 (16) (JS.4.6) 9 Ode mko wgi-sheman. odE mEko wE- gi- EshEm -a -n this bear 3- PST- feed.s.o.\TA -DIR -OBV.I This bear was feeding them. (17) (JS.4.3) 18 O negdoshas wgi-nizhokmagwan. ow negEdosha -es wE- gi- nizhokEmEw -Egwan that. AN horse -DIM 3- PST- help.s.o.\TA -3'/35.I The pony helped them. 19 E-bwamshe- je-yewawat negdoshayen e- bwamEshE - EjE- EyEw -awad negEdosha#y -En FCT- before - where- have.s.o\TA -35/3 '.C horse -OBV wgi-wbesh ' egwan seksiyen wgetganeswa. wE- EsEkEsi -yEn wE- gEtEgan -es -wa 3- deer -2.C garden -DIM -3PL Before they had the pony, the deer were ruining their gardens. (18) (JS.4.2) 47 Iw je e-gi-babgemat iw jE e- gi- babEgEm -ad and FCT- PST- DUP . strike . s . o . on . the . head\TA -3/3'. C niw ndemozeyen. niw mEdEmoze#y -En that. OBV old. woman -OBV So he knocked the old lady (in the head). 48 Jak bkwezhgasen wa-mijet zhiw jag bEkwezhEgas -En CC.wi- mij -Ed zhiw all cracker -PL CC.FUT- eat.s.t\TI -3.C there gi-tene . gi- te -nE PST- be . in . a . certain . place -O'.I All the crackers for her to eat were there. 49 Iw je e-gi-bkwenshkodwat niw iw jE e- gi- bEkwenEshkodEw -ad niw and FCT- PST- choke . s . o . \TA -3/3 '.C that . OBV 262 bkwezhgasen mine ±w ziwabo abte bEkwezhEgas -En minE iw ziwabo abEtE cracker -OBV and that.INAN cider half e-gi-zigwebnek . e- gi- zigwebEn -Eg FCT- PST- spill. s.t.\TI -3/0. C He stuffed the crackers in her mouth and poured out half of the cider. (19) (JS.4.4) 39 O mwe gi-wzam-debsenyet jo mamda ow mEwe gi- Ozam- debEsEnye -d jo mamda that. AN wolf PST- too. much- be.full\AI -3.C not possible e-wi-majnewit e- wi- mEmajEnEwi -d FCT- FUT- move.away\AI -3 . C e-pich-dbomayek . e- pij- dEbom -ayE -g FCT- so. much- talk . over . s . o . \TA -PASS -3 . C The wolf was too full; he couldn't move away while they talked over (what to do about) him. (20) (JS.4.1) 8 Iw je o wabozo zhiw gi-dbendagze odanek jo je iw jE ow wabozo#y zh iw gi- dEbEnEdagEzE odan -Eg jo jE and that. AN rabbit there PST- belong\AI . 3 . I town -LOC not but mamda 1 mamda iw possible that.INAN e-wi-zhe-nsawat mamwech bshe e- wi- EzhE- nEs -awad mamwej bEshE FCT- FUT- in. a. certain. way- kill.s.o.\TA -35/3'. C more EMPH gego gjiyek bama a-je-nsawat. gego gjiyEg bama a- EjE- nEs -awad something better later MOD- FUT- kill.s.o.\TA -35/3 '.C Since the Rabbit belonged to the village, they couldn't kill him as they please; they would have to get something more on him in order to kill him. 263 (21) (JS.4.6) 10 I je iw pi neshnalsek e-giwsewat jo iw jE iw Opi EnEshEnabe -g e- giwse -wad jo and that.INAN when person -PL FCT- hunt\AI -35. C not wgi-widpemasiwan wE- gi- wiDpem -a -si -wan 3- PST- sleep. together\TA -DIR -NEG -35/3'. I wdekweyomwan babkan zhena gi-nbek. wEdE- Ekweyom -wan babEkan EzhEna gi- nEbe/a -g 3- wife -35.0BV different EMPH PST- sleep\AI -35.1 And when people went hunting, they didn't sleep with their wives; they slept separately. (22) (MD 102594) 1 O, neko ngi-babzedwak neshnabek nEko nE- gi- DUP.bEzEdw -a -g EnEshEnabe -g oh used. to 1- PST- listen. to\TA -DIR -1/35.1 person -PL e-yayajmowat eyayenga jmowat . e- DUP.YajEmO -wad CC . yayenEga jEmO -wad FCT- tell\AI -35. C CC . laugh . about . s . t . told\AI -35. P 1 used to listen to the people telling stories; something they laughed about. 2 [Iw je] ni wabozoyen ngodek niw wabozotty -En nEgOd -Eg that.OBV rabbit -OBV one -LOC e-gi-ya jmawat . e- gi- YajEm -a -wad FCT- PST- tell. about. s.o\TA -DIR -35/3'. C Once they told about Rabbit. 3 O, bnewi neko o wabozo gi-gnewanwe. o bEnEwi nEko ow wabozo#y gi- gEnEwanwa/e oh long. ago used. to that. AN rabbit PST- have . a . long . tail\AI . I Oh, at one time. Rabbit had a long tail 4 Gi-gnewanwedek kedwik . gi- gEnEwanEwa/e -dEg EkEdO -g PST- have.a.long. tail\AI -DUB. I say\AI -35.1 He must have had a long tail, they say. 264 5 Iw je i wech-shkwanwat iw jE iw CC.wEjE- EshkwanEwa/e -d and that.INAN the . reason . why- have . a . short . tail\AI -3.C ngom ga-zhewebzet . nEgom CC.gi- EzhEwebEzE -d today PST- have . happen . to . one . a . certain . way\AI -3.C That 's why he has a short tail today, because of what happened to him. Continued. . . . 56 Iw je iw yedek iw jE iw yedEg and that.INAN must.be wech-ngom-shkwanwat CC.wEjE- nEgom- EshkwanEwa/e -d the . reason . why- today- have . a . short . tail\AI -3.C o wabozo, gi-kedwik neko gi ow wabozo#y gi- EkEdO -g nEko giw that. AN rabbit PST- say\AI -35.1 used. to those. AN gekyajek neko e-gi-wnanodogwa CC.gEkya -jEk nEko e- gi- OnanodEw -Egwa be.old\AI -35P used. to FCT- PST- hear.s.o\TA -1/35. C e-yanga jmowat . e- yanEgajEmO -wad FCT- tell. s. t. funny\AI -35. C That's must be why Rabbit has a short tail today, the elders used to say, when I heard them telling funny stories. (23) (JS.4.3) 1 Ngom wdopi wemtegozhl yewak naganit. nEgom Odopi wemEtEgOzhi EyE -wag CC.nigani -d today now French be . in . a . place\AI -3.1 lead\AI -3.P Up to today, the French are the leaders somewhere. 2 Iw je ngom wdopi nnodamen weye iw jE nEgom Odopi nE- nod -a -mEn weye and today now 1- hear.s.t\TI -DIR -15.1 someone e-wepodek biwabek wizhgya e- wepod -Eg biwabEk wizhEgya FCT- hit.s.t.\TI -3/0. C iron be. solid 265 e-nayek wi zhe ibe Kansas e- En -ayE -g wi EzhE ibE Kansas FCT- say. to. s.o\TA -PASS -3 . C EMPH EMPH there Kansas memek . memEg especially Nowadays we hear someone blacksmithing, especially there in Kansas, they say. O je ye o gche-mnedo eng[e]t wemtegozhi . ow jE ye ow gEchE- mEnEdO wemEtEgOzhi#y but . that ' s . the . one . (AN) great- spirit French That's the great spirit of the French. ye o gangezot wemtegozhi ekdonegek . ow ye ow CC.ginEgEzO -d wemEtEgOzhitty CC.EkEdO -nEgEg that ' s. the. one. (AN) be.lost\AI -3.P French say\AI -35. P That's the lost French, so they say. 1 je ngom bme-yewak zhena iw jE nEgom bEmE- EyE -wag EzhEna and today in . the . process . of- be . in . a . place\AI -3.1 EMPH nekmek . EnEkEmEg different .places Now he is moving around in different places. Jo win gdemagzesi ginan wi jo win gEdEmagEzE -si ginan wi not 3. EMPH be.poor\AI -NEG.I 12. EMPH EMPH eneshnabewigo gdekdomen . e- EnEshEnabewE yEgO gEd- EkEdO -mEn FCT- be.IndianX AI 12. C 2- say\AI -15.1 He is not poor; we who are Indians say that. Wemtegozhi maneton wzaw-zhonya wemEtEgOzhi manet -o -n wEzaw- zhonya French have . plenty . of . s . t . \TI -OBJ -OBJ. I yellow- money mine mkede-biwabek . minE mEkEde- biwabEk and black- iron The French have lots of gold and black iron. 266 8 Mine ngom e-gkendemgo bgoch-negdoshayek mine minE nEgom e- gEkendEm -Ego bEgOj- negOdosha#y -Eg minE and today FCT- know\AI2 -12. C wild- horse -PL and seksik jak zhena e-yemgek. EsEkEsi -g jag EzhEna e- EyE -mEgEg deer -LOG all EMPH FCT- be . in . a . place\AI -AUG.O.C And today we know wild horses and deer and so forth are there. 9 Ode je nene win wdebenman. odE jE EnEnE#w win wE- dEbenEm -a -n this but man 3 . EMPH 3- own.s.o\TA -DIR -OBV.I This man owns them. 10 Ode je wemtegozhi gzhe-mnedon odE jE wemEtEgOzhi gEzhe- mEnEdO -n this but French great- spirit -OBV wgi-nizhokmagon e-wi-mishgwezet ode wE- gi- nizhokEmEw -agOn e- wi- mishEgwEzE -d odE 3- PST- help.s.o\TA -0/3. C FCT- FUT- be . powerful\AI -3 . C this je anet gikansenan Spanish jE anEd gE- # ikanEs -Enan Spanish but some 2- brother -12.P0SS Spanish e-nayek ode e- En -ayE -g odE FCT- say. to. s.o\TA -PASS -3 . C this wdekwenzhgewan . wEdE- EkwenzhEgEw -a -n 3- beat . s . o. in. a. contest\TA -DIR -OBV.I Now God helped the French to be powerful, but his brother the Spanish was victorious, they say. 11 I je ngom wdopi ode wemtegozhi nwech zhe ninweze iw jE nEgom Odopi odE wemEtEgOzhi nwEj zh E ninwEzE and today now this French more EMPH be . weak\AI . 3 . I zhode kik . ZhodE EkE -g here earth -LOC Up to today, the French are very weak in the world. 12 Ngodek je ode wemtegozhi wgi-nizhokmowen nEgOd -Eg jE odE wemEtEgOzhi wE- gi- nizhokEmEw -En one -LOC but this French 3- PST- help.s.o.\TA -OBV.I 267 neshnaben . EnEshEnabe -n Indian -OBV At one time, the French helped out the Indians. 13 I je i pi ode wemtegozhi iw jE iw pi odE wemEtEgOzhi and that's. when this French wgi-minan ngemwen wE- gi- miN -a -n nEgEmOwEn 3- PST- give. to. s . o\TA -DIR -OBV. I song i je ye i ngoiti gode neshnabek iw je ye iw nEgom godE EnEshEnabe -g that ' s . the. one. (INAN) today these. AN Indian -PL e-yewat i je ode ngom nim'ediwen e- EyE -wad iw jE odE nEgom nimE ' EdiwEn FCT- be . in . a . place\AI -35. C and this today dance gode neshnabek e-gche-yowat . godE EnEshEnabe -g e- gEchE- EyEw -Ewad these. AN Indian -PL FCT- really- use.s.t.\TI -35/1. C At that time the French gave him a song, and that's the one these Indians here use in their dancing to this day. (24) (JS.4.3) 14 I je o wemtegozhi e-gi-nat iw jE ow wemEtEgOzhi e- gi- En -ad and that. AN French FCT- PST- say . to . s . o\TA -3/3'. C niw gigabeyen "Nasena niw gigabe#y -yEn nasEna that. OBV boy -2C be. careful zhechgen ezh-widmonan . " EzhEchEge -n CC.EzhE- widEmEw -nan do. things . a. certain. way\AI -OBV thus- tell.s.o\TA -1/2. C The French told one boy, "Be careful to do things the way I tell you to. " 268 15 I je o wemtegozhi e-wishteyaywat iw jE ow wemEtEgOzhi e- wishEtEyay -Ewad and that. AN French FCT- blacksmith\AI -35. C e-gkeno 'mewat ni gigabeyen. e- gEkEno ' EmEw -ad niw gigabe#y -yEn FCT- teach. s.o\TA -3/3'. C that . OBV boy -2 . C So that French (Spirit) was teaching the boy how to blacksmith. 16 I je o gigalse wikapi iw jE ow gigabe#y wikapi and that. AN boy finally e-gi-ne-wishteyaywat e- gi- nE- wishEtEyay -Ewad FCT- PST- start. to- blacksmith\AI -35. C e-gi-gkeno 'mowat niw wmeshomsen. e- gi- gEkEno 'EmEw -ad niw wE- mEshomEs -En FCT- PST- teach. s.o\TA -3/3'. C that . OBV 3- grandfather -OBV Finally, the boy started to blacksmith, and he taught his grandfather. n Wikapi e-gi-negdoshayensawat mine zhena wikapi e- gi- negOdoshayEnEs -awad minE EzhEna finally FCT- PST- have . horses\TA -35/3 '.C and EMPH gego . gego something Finally, they had a pony and so forth. 18 O negdoshas wgi-nizhokmagwan. ow negEdosha -es wE- gi- nizhokEmEw -Egwan that. AN horse -DIM 3- PST- help.s.o.\TA -3'/35 The pony helped them. 19 E-bwamshe- je-yewawat negdoshayen e- bwamEshE - EjE- EyEw -awad negEdosha#y -En FCT- before - where- have.s.o\TA -35/3 '.C horse -OBV wgi-wbesh ' egwan seksiyen wgetganeswa. wE- EsEkEsi -yEn wE- gEtEgan -es -wa 3- deer -2.C garden -DIM -3PL Before they had the pony, the deer were ruining their gardens. 269 20 Gigabe e-ggenowat ni gigabe#y e- gEgEnow -ad niw boy FCT- talk. to. s.o. \TA -3/3'. C that . OBV negdoshayen, "Ni je wa-zhechgeyan?" negOdosha#y -En ni jE CC.wi- EzhEchEge -yan horse -OBV what FUT- do . things . a . certain . way\AI -l.C The boy asked the pony, " What should I do?" 21 I je o negdosha e-nat, iw jE ow negEdosha#y e- En -ad and that. AN horse FCT- say . to . s . o\TA -3/3'. C "Wigbish mtegok wdenen wigObish mEtEg#0 -g OdEn -En basswood. bark tree -LOG get . s . t . from. someplace\AIO -2 . IMP ge-dkobdon nkwegnak gekwedso gE- dEkobEd -on n- # kwegEn#a -g EgEkwEdEso FUT- tie.s.t.\TI2 -3/0. IMP 1- neck -LOG so .many . times egme-kezhyep ge-giwta ' omgon GG.EgEmE- gEgEzhyeb gE- giwta ' omEgo -n every- morning FUT- ride . around. in . a . circle\AI -2 . IMP iw ggetganwa . " iw gE- gEtEgan -wa that.INAN 2- garden -2PL And the pony said, "Get some bark, tie it around my neck, and ride me around your garden every morning. " 22 I je gi seksik e-wi-zegzewat . iw jE giw EsEkEsi -g e- wi- zegEzE -wad and those. AN deer -PL FGT- FUT- be.scared\AI -35. G The deer will be scared. 23 Nesh je gego zhe nEzh jE gego zhE contrarily but something EMPH gwi-zhe-ngok . " gE- wi- EzhE- En -Ego -g 2- FUT- in. a. certain. way- say . to . s . o\TA -INV -35.1 Of course, they will say something to you. " 270 2i O seksi e-kedot "Wegni je o ow EsEkEsi e- EkEdO -d wegni jE ow that. AN deer FCT- say\AI -3.C what that. AN wakayabde wakayabEde have . a . round . tooth bye-zizdeyatek?" bye- zizEdEyatEg come- have . s . t . sticking . out . between . one ' s . legs\AI .3.1 The deer said "What does that round-tooth have sticking out between his legs?" 25 Egme-kezhyep zhena o je wemtegozhi CC.EgEmE- gEgEzheb EzhEna ow jE wemEtEgOzhi#y every .morning EMPH that. AN but French nizhokmowen i je mine wa-mijwat nizhokEmEw -En iw jE minE CC.wi- mij -Ewad help.s.o.\TA -3/3 '.I and and FUT- eat.s.t\TI -3/3 '.P wiyas o gi-wje-wdetnanawa. wiyas ow gi- wEjE- OdEdEn -a -nawa meat that. AN PST- how- get.s.o.\TA -DIR -35/0.1 Every morning the French Spirit helped them, and that's how they obtained their meat to eat. 26 Ga-gish- jagnenet CC.gi- gizh- jagEne -nEd CC.PST- finish- be.dead\AI -3'.C wdenwemagnen wmeshomsen wEdE- nwemagEn -En wE- mEshomEs -En 3- relative -OBV 3- grandfather -OBV ga-gish-mbonet e-gi-majit. CC.gi- gizh- nEbo -nEd e- gi- maji -d PST- finish- die\AI -3'.C FCT- PST- leave\AI -3 . C After his relatives and grandfather died, he left. 271 27 I je geyabe wemtegozhi nizhokmowan geyabe iw jE geyabE wemEtEgOzhi nizhkEmEw -o -wan geyabE and still French help.s.o\TA -OBJ -35/3 '.I still je ngom gnizhokmagnan. jE nEgom gE- nizhkEmEw -EgO -Enan but today 2- help.s.o\TA -INV -3/12.1 Still the French helped him, and is helping us to this day. 28 I je ngom pi neshnabe wemtegozhi mskwe iw jE nEgom Opi EnEshEnabe wemEtEgOzhi mEskwe and today when Indian French blood we j-gwgezhkek o wemtegozhi CC.wEjE- gOgEzhk -gEg ow wemEtEgOzhi#y CC.why- have. s . t . in. one ' s .body\AI -35/0. P that. AN French e-gi-zhwenmat . e- gi- EzhEwenEm -ad FCT- PST- bless. s.o. \TA -3/3'. C Up to today Indians have French blood inside them, because the French (Spirit) blessed them. 29 Ode gigabe e-gi-majit e-gi-byat odE gigabe#y e- gi- maji -d e- gi- bya/e -d this boy FCT- PST- leave\AI -3 . C FCT- PST- come\AI -3 . C odanek neshnaben eyenet, odan -Eg EnEshEnabe -n CC.EyE -nEd town -LOC Indian -OBV CC . be . in . a . place\AI -3'.P ga-gkendek je ni wemtegozhiyen CC.gi- gEkend -Eg jE niw wemEtEgOzhi -yEn PST- know.s.t\TI but that . OBV French -2.C wgi-gkeno 'mowan neshnaben wE- gi- gEkEno'EmEw -a -n EnEshEnabe -n 3- PST- teach. s.o\TA -DIR -OBV person -OBV wa-zhi ' enet . CC.wi- Ezhi' -EnEd CC.FUT- be.in.a.certain.place\AI -3'.C This boy left and came to where there was an Indian village; what he learned from the French he taught the people who were there. 30 Ga-gish-gkeno 'mowat wiznaben CC.gi- gizh- gEkEno'EmEw -ad wiznabe -n CC.PST- finish- teach. s.o\TA -3/3'. C fellow. men -OBV 272 wa - z he tone t , gego CC.wi- OzhEto -nEd gego FUT- make.s.t.\TI -3'.C something wgi-nan: wE- gi- En -a -n 3- PST- say. to. s.o\TA -DIR -OBV " Gego nsedi ' kegon , " gegO nEsEdi' -kegon don't kill . one. another\AI -25. IMP wgi-nan. wE- gi- En -a -n 3- PST- say. to. s.o\TA -DIR -OBV After he taught his fellow people what to do, he told them something: "Don't kill one another, " he said. 31 I je o wemtegozhl e-gi-nat iw jE ow wemEtEgOzhi e- gi- En -ad and that. AN French FCT- PST- say . to . s . o\TA -3/3'. C niw gigabeyen ga-widmowak niw gigabe#y -yEn CC.gi- widEmEw -wag that. OBV boy -2.C PST- tell . s . o\TA -3.1 e-wi-bwa-mje-dodadwat, mine i je ngom wdopi e- wi- bwa- mEjE- dodad -Ewad minE iw jE nEgom OdE Opi FCT- FUT- NEG- bad- do . to . e . o . \AI -35. C and and today now neshnabek enwek-dbenbwek . EnEshEnabe -g enwek- dEbEnEbwe -g Indian -PL better- be . civilized\AI -35.1 And the French told the boy what to tell them, that they should not abuse each other, and so up to this day, the Indians are surely civilized. (25) (JS.4.4) 32 I je ngom wdopi dewe ' gen-nim ' ediwen iw jE nEgom Odopi dewe'EgEn - nimE'EdiwEn and today now drum - dance debwetanawa neshnabek debwet -a -nawa EnEshEnabe -g believe. s.t\TI -OBJ -35/0.1 Indian -PL 273 i ye i gagowat ni wemtegozhiyen . iw ye iw ? niw wemEtEgOzhi -yEn that ' s. the. one. (INAN) ? that.OBV French -2C Up to this day the Indians beUeve in the drum dance; that's the one the French told him about. 33 Iw je ngom wdopi jak neshnabek iw jE nEgom Odopi jag EnEshEnabe -g and today now all Indian -PL wdebwetanawa ode madmowen iw je wE- debwet -a -nawa odE madEmowEn iw jE 3- believe. s.t\TI -OBJ -35/01 this belief and we j -mno-widokwdadwat . CC.wEjE- mEnO- widokwdad -Ewad why- good- be . f riends\AI -35. C And up to today, all the Indians believe this way and that's why they are good friends. 34 Nchiwenmok e-wabdawat ngom wdopi . nEchiwenEmO -g e- wabEd -awad nEgom Odopi be.glad\AI -35.1 FCT- see.s.t\TI -35/3'. C today now They are happy to see each other up to today. 35 Iw je ekwak ode wemtegozhi yajmowen. iw jE CC.Ekwa -g odE wemEtEgOzhi YajEmowEn and be.long\II -O.C this French story So that's how long this French story is. (26) (MD 102694) 12 I je ge wi zhi o gagtanago i yedek . iw jE ge wi zhiw ow gagEtanago#y iw yedEg and also EMPH there that. AN crocodile that. INAN must.be So must be Crocodile was there. 13 Beshoch zhena zhi jigbyek [ge] beshoj EzhEna zhiw jigbyeg near EMPH there by. the. water e-gegwi jek . e- DUP.gwiD -Eg FCT- float. in. the. water\AI -3.C He was floating in the water near the shore. 274 13 Zagwjanegwi jen zhi zagwjanegwiD -En zhiw have . one ' s . nose . float\AI -3.1 there His nose was sticking out there. (27) (MD 102694) 46 [win] ibe shkweyak gi-nshkweshen [win] ibE Eshkweyak gi- nEshkweshEnO [3.EMPH] there last PST- lie . in . last . place\AI . 3 . I i ga-nakwneget gagtanago . iw CC.gi- nakOnEge -d gagEtanago#y that.INAN CC.PST- plan . things\AI -3 . P crocodile The Crocodile that planned it lay at the end, there in last place. (28) (JS.4.4) 28 Espen o mtegok gi-gdegosi esEpEn ow mEtEg#0 -g gi- gEdEgosi raccoon that. AN tree -PL PST- climb . up\AI . I e-wawabmat niw mwen wete e- DUP.wabEm -ad niw mEwe -n wEtE FCT- watch. s.o\TA -3/3'. C that . OBV wolf -OBV for. sure zhe e-gi-bdek ' egaznet . zhE e- gi- bEdEk' -EgazO -nEd EMPH FCT- PST- sting . s . o . \TA -PASS -3'.C The Raccoon was high (in a tree) and saw the Wolf get badly stung. 29 I je o mwe jo gi-nshkadzesi ; neshnege iw jE ow mEwe jo gi- nEshkadEzE -Esi neshEnEge and that. AN wolf not PST- be.angry\AI -NEG.I all. the. same megwa gi-dnendan megwa gi- EdEnend -a -n still PST- think. that . s . t. is . in. a. certain. place\TI -OBJ -OBJ. I i wa-zhyawat iw CC.wi- Ezhya/e -wad that.INAN FUT- go.there\AI -35. C 275 e-wi-gmodwat gokoshen. e- wi- gEmOd -wad gokosh -En FCT- FUT- steal\AIO -35. C pig -OBV That Wolf didn't get mad; he still thought the meat would be there, and wanted to go there and steal that pork. (29) (JS.4.4) 5 Gete zhena gEtE EzhEna for. sure EMPH e-gi-nkweshkwat mwen. e- gi- nEkweshkEw -a -d mEwe -n FCT- PST- meet.s.o.\TA -DIR -3.1 wolf -OBV Sure enough, he (the Raccoon) met Wolf. 6 "Nshi, gde-ton ne nshi gEdE- Et -o -n nE little. brother 2- have.s.t\TI -OBJ -OBJ. I Q gego wa-mijyan?" gego CC.wi- mij -yan something FUT- eat.s.t\TI -1/0. C e-nat espenen. e- En -ad esEpEn -En FCT- say. to. s.o\TA -3/3'. C raccoon -OBV "Brother, do you have anything to eat? " he said to the Raccoon. 7 "Jo zhe kwech bkeji nde-ton jo zhE kwej bEkeji nEdE- t -o -n not EMPH hardly a. little 1- have.s.t\TI -OBJ -OBJ. I wa-mijyan nawkwek," CC.wi- mij -yan nawEkweg FUT- eat.s.t\TI -1/0. C at . noon e-nat espen. e- En -ad esEpEn FCT- say. to. s . o\TA -3/3'. C raccoon "Not much, I just have a little for my own dinner, " said the Raccoon. 8 Mwe e-natewat, "Wegni je mEwe e- natEw -ad wegni jE wolf FCT- ask.s.o.\TA -3/3'. C what 276 etoyen? " CC.Et -o -yEn have.s.t\TI -OBJ -2 . C Wolf asked him, "What do you have ? " 9 Espen e-nat, "Mteno zhe na esEpEn e- En -ad mEtEno zhE na raccoon FCT- say . to . s . o\TA -3/3'. C only EMPH EMPH bkeji gokosh-wzhey ndesa," bEkeji gokosh-w#zhEy nEd- Es -a a. little pork. rind 1- have.s.o\TA -DIR.I e-nat. e- En -ad FCT- say. to. s.o\TA -3/3'. C Raccoon said, "I have just a httle meat-rind, " he said. 10 Mwe e-nat, "Mojma mEwe e- En -ad mojma wolf FCT- say. to. s.o\TA -3/3'. C please shemshen o wzhey. " EshEm -shEn ow w- #zhEy feed.s.o.\TA -2/1. IMP that. AN 3- skin Wolf said, "Please feed me that rind. " 11 I je o espen iw jE ow esEpEn and that. AN raccoon msach e-gi-minat. msaj e- gi- miN -ad finally FCT- PST- give . to . s . o\TA -3/3 '.C So the Raccoon finally gave it to him. (30) (JS.4.2) 62 Ga-gish-ngo ' wawat gigabe neyap CC.gi- gizh- nEgO'Ewa -wad gigabe#y neyab CC.PST- finish- bury.s.o.\TA -35/3'. C boy back e-wawidmewat niw keweziyen, e- DUP.widEmEw -Ewad niw kEwezi#y -En FCT- tell. s. t. to. s.o.excitedly\TA -3/3'. C that . OBV old. man -OBV 277 "Nmesho, ngodwak gwkengo'gazo nE- mEsho nEgOdwag gOkenEgo ' Egazo 1- grandfather .VOC one . hundred be . buried. with . s . o . \AI . I o ndemoze . " ow mEdEmoze that. AN old. woman After they buried her, the boy went back and excitedly told the old man, "Grandfather, one hundred dollars is buried with that old lady. 63 Nwi -mon ' wa . " nE- wi- mon ' Ew -a 1- FUT- dig. s.o.up\TA -DIR.I I'm going to dig her up. " 64 Kewezi "Jo, gego" wdenan . kEwezi#y jo gegO wEd- En -a -n old. man no don't 3- say . to . s . o\TA -DIR -OBV.I But the old man said, "No, don't. " 65 "Gda-bon-gdemagzemen iw ngodwak," gE- da- bon- gEdEmagEzE -mEn iw nEgOdwag 2- MOD- quit- be.pitiful\AI -15.1 that.INAN one . hundred wdenan . wEd- En -a -n 3- say. to. s.o\TA -DIR -OBV "We could quit living poorly with that hundred, " he said to him. 66 Kewezi "Gego" wdenan. kEwezitty gegO wEd- En -a -n old. man don't 3- say . to . s . o\TA -DIR -OBV.I "Don't, " said the old man. 278 (31) (MD 102694) 27 "A, iw zhe yedek e-wi-dkemozh ' ewat a iw zhE yedEg e- wi- dEkEmozhE' -Ewad ah that.INAN EMPH must.be FCT- FUT- take . s . o . across\TA -35/1. C gode," zhede'e o wabozo. godE EzhEde'a/e ow wabozo#y these. AN think\AI.3I. that. AN rabbit "Ah, must he they will take me across, " thinks the Rabbit. (32) (MD 102694) 25 "Gego zhe ode gagtanago nwi-nakwnek, " gego zhE odE gagEtanago#y nE- wi- nakwEn -EgO something EMPH this crocodile 1- FUT- plan . for . s . o\TA -INV.I e-zhde'at o wabozo. e- EzhEde'a/e -Ed ow wabozo#y FCT- think\AI -3.C that. AN rabbit "This Crocodile has something planned for me, " thought the Rabbit. (33) (MD102694) 36 "O, wzam ne zhe gete ode? o Ozam nE zhE getE odE oh too. much Q EMPH for. sure this 37 Gagtanago nwe jitmagodek?" e-zhde'at. gagEtanago#y nE- wEjitEm -agO -EdEg e- EzhEde'a/e -Ed crocodile 1- help.s.o\TA -PASS -DUB.l.I FCT- think\AI -3.C "Oh, can this really be? Will Crocodile really help me? " he thought. (34) (MD 102694) 15 "A! Nshi! Ni je a nshi ni jE ah little . brother what ezhwebzeyen? " CC.EzhEwebEzE -yEn have . happen . to . one . a . certain . way\AI -2 . C wdenan ni wabozoyen. wEd- En -a -En niw wabozo#y -En 3- say. to. s.o\TA -DIR -3/3'. I that . OBV rabbit -OBV "Ah, little brother! What's the matter?" he said to the Rabbit. (MD 1 02 694. 01 5) 279 (35) (MD 102694) 19 "O, jo wi zhe na gego abje o jo wi zhE na gego abEjE oh not EMPH EMPH EMPH something very yawsenon i , " kedo o yaw -sEnon iw EkEdO ow be. a. certain. thingXlI -NEG.O.I that.INAN say\AI.3.I that. AN gagtanago . gagEtanago#y crocodile "Oh, there 's nothing much to that, " said the Crocodile. (36) (see 30) (37) (MD 102694) 48 O, [nme pa zho] megwa e-gche-bmebtot bama zhe o megwa e- gEchE- bEmEbEto -d bama EzhE oh still FCT- really- run\AI -3.C later EMPH gete. . . [o] bikwa zhe na wangoyane getE bikwa zhE na wanEgoyanE for. sure just. like EMPH EMPH be . an . underwater . hole\II . . I wiye gego e-wabdek. weye gego e- wabEd -Eg someone something FCT- see.s.t\TI -3/0. C Oh, as he was dashing across, he soon [saw something] that looked just like a hole. (38) (MD 102694) 59 Iw je o wabozo neko iw jE ow wabozo#y nEko and that. AN rabbit used. to e-wi-wabmek megwa e- wi- wabEm -Eg megwa FCT- FUT- see.s.o\TA -1/3. C still e-peno jewyan, iw zhe neko i e- EpEnojew -yan iw zhE nEko iw FCT- be.a.child\AI -l.C that.INAN EMPH used. to that.INAN 280 e-gwdenmewek iw wzewangos e- gOdEnEmEw -Eg iw wE- zEwanEg#o -s FCT- try. feeling. s.o\TA -1/3. C that.INAN 3- tail -DIM And when I used to see the rabbit, when I still was a child, I used to feel his little tail. 60 O, gete zhe na ode gi-gishk jegadenek iw o getE EzhE na odE gi- gishkE jEgade -En -Eg iw oh for. sure EMPH EMPH this PST- be.spared\II -OBV -35.1 that.INAN wzewangos, neko ngi-zhde'a. wE- zEwanEg#o -es nEko nE- gi- EzhEde'a/e 3- tail -DIM used. to 1- PST- think\AI . I Oh, for sure that little tail was bitten off, I used to think. 61 Nmet se na yedek w± na! nEmED sE na yedEg wi na I. don't. know must.be EMPH EMPH / don't know about that! (39) (JS.4.6) 8 I je ode nene e-gi-nme-ninwezet iw jE odE EnEnE#w e- gi- nEmE- ninEwEzE -d and this man FCT- PST- getting.to.be- be.weak\AI -3C e-wzam-bkedet i je ode kwe mine o e- Ozam- bEkEde -d iw jE odE Ekwe minE ow FCT- too. much- be . hungry\AI -3.C and this woman and that. AN gigabe pene zhena winwa gi-gimoch-wisnik . gigabe#y pEne EzhEna winEwa gi- gimoj- wisEnE -g boy always EMPH 35. EMPH PST- secretly- eat\AI -35.1 So this man got to be weak from hunger, but the woman and the boy were secretly eating. (40) (JS.4.2) 116 I je gi wewiwdeyek e-gi-yewat iw jE giw DUP.wiwdEyEg e- gi- EyE -wad and those. AN married . couple FCT- PST- be . in . a . place\AI -35. C jayek debendemwat e-gi-mbomgek jayeg CC.dEbendEm -wad e- gi- nEbO -mEgEg all CC.own\AI -35. P FCT- PST- die\AI -AUG.O.C 281 e-gi-gdemagzewat . e- gi- gEdEmagEzE -wad FCT- PST- be.poor\AI -35. C And the couple settled, and all that they owned [their stock and fowl] died, and they were poor. 117 Mine wzhonyamwa e-gi-jagsanek. minE wE- zhonya -Em -wa e- gi- jagEsa -n -Eg and 3- money -POSS -3PL FCT- PST- run . out . of . s . t . \AIO -OBJ -3C Also their money ran out. 118 O je kewezi mine gigabe gi-mno-bmadzik . ow jE kEwezi#y minE gigabe#y gi- mEnO- bEmadEzE -g that. AN but old. man and boy PST- good- live\AI -351 But the old man and the boy lived happily. (FN18)(JS.4.1) 20 E-pitajmewat ngot mine e-kedot, e- EpitajEm -Ewad nEgOd minE e- EkEdO -d FCT- talk. a. certain. amount\AI -35C one and FCT- say\AI -3C "Shebzhi ngi-nek, 'Nin mEshEbEzhi nE- gi- En -EgO nin tiger 1- PST- say . to . s . o\TA -INV I . EMPH nda-nsa, ' kedo . " nE- da- nEs -a EkEdO 1- MOD- kill.s.o.\TA -DIR say\AI . 31 While they were talking, another man said, "Lion said to me 7 can kill him '. " 2 Example from Chapter 7 (13) (MD 102694) 6 Jigbyek ibe e-pa-zhyat. jigbyeg ibE e- bEba- Ezhya/e -d by. the. water there FCT- around- go.there\AI -3.C 7 "O, begesh na ezhi gameyek o begEzh na ezhi gameyEg oh would. that EMPH over. there across . the . river 282 gshketoyan gEshEkEt -o be. able. to.do. s . t\TI -OBJ e -by ay an, " -yan e- bya/e -yan -IC FCT- come\AI -l.C e-kedot . e - EkEdO -d FCT- say\AI -3 . C E -dnednangedok e- EdEnEdEnanEgEdO -Eg FCT- be . the . sound. of . s . o . talking . in . a . certain . place\II -O.C j igbyek . jigbyeg by. the . water Gegpi zhe gegEpi zhE finally EMPH gwagwashkze ' o . gwagwashkEzE ' o DUP . j ump . up . and . down\AI .3.1 He went around there by the water. "Oh, I wish I was able to get across over to there, " he said. He was talking to himself along the river. Finally, he started jumping up and down. 283 Appendix C. Crane Boy Narrative I me se ngodek jejakok that.INAN EMPH EMPH one=LOC crane=PL e-gche-wzhenwiwat e-nme-dgwagek FCT-really-get. ready\AI=35C FCT-getting . to . be-be . fall\II=OC wech-gzhatek CC . towards -be . hot . weather\II=OP Once when it was getting close to Autumn, cranes were preparing for spending the winter in the south; at first, they talked to each other about when they would start, as was customary. to 00 e-we-bbonshewat , FCT-go . and- spend. the .winter.in.a.certain.place=35C netem zhe na e-widmedwat o pi first EMPH EMPH FCT-say . to . e . o . \AI = 35C that. AN when wa-majiwat neko . CC. FUT-leave\AI=35P used. to Iw je i e-dwagnekewat and that.INAN FCT-store . things . away\AI=35C wa-mijwat e-pich-bmodegzewat . CC. FUT-eat. s.t\TI=35/0P FCT-while-move\AI=35C They stored things away to eat while they moved. Nangodek nyew gon dnekiwdek Sometimes four day happen. in. a certain . place\II=DUB . 01 e-wzhenwiwat e-nwepwankewat wa-mijwat. FCT-get. ready\AI=35C FCT-pack . a . lunch\AI=35C CC . FUT-eat. s.t\TI=35C Sometimes it must have taken four days for them to get ready, packing their food to eat. Iw se ga-gish-wzhenwiwat that.INAN EMPH CC . PST-f inish-get . ready\AI=35C e-wi-majiwat zhye bos-kezhyep. FCT-FUT-leave\AI=35C EMPH very-early So when they were finished getting ready, they would leave very early. E-bkonyak e-gche-giwnezwat FCT-be . night\II=OC FCT-really-f ool . around\AI=35C gigabesek e-pich-nchiwenmowat boy=DIM=PL FCT-so .much-be . glad\AI=35C At night, the boys really fooled around they were so glad to move, and one boy broke his arm. to 00 e-wi-bmodegzewat, iw o ngot gigabe FCT-FUT-move\AI=35C that.INAN that. AN one boy e-gi-boknekeshkwayek . FCT-PST-break . one' s . arm\AI=PASS=3C Iw se niw wmezodanen that.INAN EMPH that . OBl^ 3-parent=OBl^ e-wi-ngengot FCT-FUT-leave . s . o . behind\TA=3' /3C gbe-bbon e-got. through. all . of -winter FCT-say. to. s . o. \TA=3'/3C So his parents told him they were going to leave him behind all winter . An we e - gche -mwe t . although FCT-really-cry\AI=3C He really cried hard, though. "Ni je bzhe a-napnennak, anwe what EMPH MOD-deal . with . s . o . thus\TA=15/2C although bzhe gche-mwin gbe-dbek," EMPH really-cry\AI=2C through . all . of-night e-got ni wmezodanen, FCT-say. to. s.o. \TA=3'/3C that . OBV 3-parent=OBI^ "What will we do with you if you cry all night? " his parents said to him, "You did this to yourself; you don't listen to what you are told, you are too naughty. " "gin je ge zhe na gde-dodan 2. EMPH but also EMPH EMPH 2-do . something . to . s . t . \TI=OBJ=OBJ giyow, e-bwa-bzedageyen yourself FCT-NEG-listen . to . people\AI=2C to 00 e-keno 'megoyen anwe i FCT-teach. s.o. \TA=PASS=2C all . right that . INAN e-wzam-kebadzin. " FCT- too .much-be . naughty\AI=2C 9. 10. Gete zhe na gbe-dbek e-mwet gigabe . for.sureEMPH EMPH through . all . of-night FCT-cry\AI=3C boy Iw se kezhyep e-yawek, that. INAN EMPH morning FCT-be . a . certain . thing\II=OC e-wi-majiwat zhye. FCT-FUT-leave\AI=35C EMPH Sure enough, the boy cried all night So morning came, and they were leaving. J 2 Iw se gigabe e-towayek that.INAN EMPH boy FCT-put . s . t . \TI=OBJ=PASS=0C ngot wabozoyen zhiw wdeshkmodak one rabbit=OBl^ there 3-sack=L0C wa-mwajen gbe-bbon. CC. FUT-eat. s.o. \TA=DIR=3/3 ' P through . all . of-winter So one rabbit was put in his sack for him to eat for the entire winter. \2 E-bwa-wdapnat anwe, FCT-NEG-pick. s.t.up\TI=3/0C although e-gi-gzekeyewnedwat gi jejakok gyetnam FCT-PST-fly.up\AI = 35C those. AN crane=PL for. sure He didn't take it though. As the cranes flew up, the boy jumped and jumped, [trying to follow them]. to 00 <1 zhe gigabe e-gche-gwagwashkze ' ot . EMPH boy FCT-really-DUP . jump . up . and. down\AI . 3I=3C J 3 Jo mamda e-wi-gzekat; i not possible FCT-FUT-f ly . away\AI=3C that.INAN e-boknekwat . FCT-have . a . broken . arm\AI=3C 24 Iw se ga-bondemwet that.INAN EMPH CC . PST-stop . crying\AI=3C e-bme-nanibwet zhiw FCT-in . the .process . of -stand. up\AI=3C there He couldn't fly away; his arm was broken. After he stopped crying, he stood around there by the ocean. j ik -gchegem . next . to-big . lake J 5 I wi ge wi mdemoze that.INAN EMPH also EMPH old. woman e-bba-nanibwet, i je weye FCT-go . around-stand. up\AI=3C and someone So an old woman was standing around and heard something; Uke someone crying, she thought. e-nodwat e-mwenet gechwa FCT-hear. s. t\TI=3/3'C FCT-cry\AI=3' C like e-zhede ' at. FCT-think\AI=3C to 00 00 Jg "Weni je yedek a-yawet?" who must.be MOD-be . a . certain . thing\AI=3C e-kedot FCT-say\AI=3C e-bme-kenondezot, "Na se wi na FCT-in. the. process. of-talk. to. s.o. \TA=REFL=3C EMPH EMPH EMPH EMPH "Who could it be?'" she said, talking to herself, "Maybe I will be blessed by a child, " she said. nda-ne-zhwendagwes penoje 1-MOD- start. to-be . blessed\AI child e-kedot. " FCT-say\AI=3C J7 Iw je gete e-gi-naskwat and for. sure FCT-PST-approach . s . o . \TA=3/3' C ednwe wegz ene t FCT- sound. comes . from. a . certain . place\AI=3' C So she went to where the sound of him was coming from; straight across there. dbaze e-gi-zhyat. straight . across FCT-PST-go . there\AI=3C Jg Ibe e-byat jik-gchegem bama zhe na there FCT-come\AI=3C next . to-big . lake soon EMPH EMPH gete gigabeyen e-nemsenet; "Ni je for. sure boy=OBl^ FCT-walk . of f \AI=3' C what bem-zhewebzin gigabe?" along-be .in. a. certain. state\AI=2C boy e-nat, e-gi-yatwasat FCT-say. to. s.o. \TA=3/3'C FCT-PST-f all . on . back\AI=3C gigabe e-pich-zegzet . boy FCT-so .much-be . scared\AI=3C She came to the big lake there, and soon the boy had started to walk ojf; "What's the matter, boy, " she said, and the boy was so scared, he fell back. to 00 J 9 "Jo zhe na gego no EMPH EMPH something ngi -ngedgamgok 1-PST-leave. s . o . behind. with . s. t. \TA=INV=35/1I "My parents left me behind with nothing; they won 't come back until springtime, " said the boy. nmezodanek; bama nokmek wi-byek," l-parent=PL later be . springtime\II=OC FUT-come\AI=35I e-kedot gigabe. FCT-say\AI=3C boy 20 "Gego myanendeken, nin gbe-bbon don't be.sad=2PR0H I . EMPH through . all . of-winter gge-mik jewit; o gge-ndasgo; noses 2-FUT-work\AI=3C that. AN 2 . FUT-be . good. company\AI . I my . grandchild je ggeyaw," e-kedot mdemoze. but 2-FUT-be. a. certain. thing\AI . I FCT-say\AI=3C old. woman "Don't be sad, all through winter you shall work for me; you will be good company; you 'II be my grandchild, " said the old woman. 2\ Gete se e-mnewendek gigabe for. sure EMPH FCT-be . glad\AI2=3C boy e-wi-tot FCT-FUT-have. s . t\TI=OBJ=3/0 (5) C The boy was very happy to have a place where he would spend the winter. wa- je-bbonshet . CC . FUT-whe re- spend. the .winter. in. a. cert ain.place=3C to o 22 Iw se e-gi-wijewat ni that.INAN EMPH FCT-PST-go . with . s . o . \TA= 3/3' C that . OBl^ okmesen bama zhe na e-byawat wigwam grandmother=OBI^ later EMPH EMPH FCT-come\AI=35C house ga-tenek pekyegan CC . PST-be . in . a . certain . place=OBV"=0C mat . house e-je-dat o mdemoze. FCT-where-live . in . a . certain . place\AI=3C that. AN old. woman So he went with his grandmother, and soon they came to where her house was; the old woman lived in a mat-house. 23 Jak zhe na gego neshnabe-zhechgewen, all EMPH EMPH something Indian=do . things . a . certain . way\AI -NOM e-wabdek o gigalse naknen FCT-see. s. t\TI=3/0C that. AN boy mat=PL e-wenek e-zheweksek gawta-yegwan . FCT-be. good\II=OC FCT-lie . spread. out\II=OC around. something Everything was all done the Indian way; the boy saw that the mats were good, and spread out all around. 24 Iw se e-yayajmo ' got that.INAN EMPH FCT-tell . stories . to . s . o . \TA=3' /3C e-bkonyak ni okmesen. FCT-be. night\II = OC that.OBl^ grandmother-OBl^ So his grandmother told him stories at night. 25 Wa-nme-zhewebzet CC . FUT-in . the . process . of-have . happen . to . one . a . certain . way\AI=3C e-widmagot, FCT-tell. s.o\TA=3'/3C She told him what would happen, and afterward, she told him, "there is one other boy I take care of. " e-nme-gizha jmo ' got FCT- in . the . process . of- finish . telling . stories. to. s.o. \TA=3 '/3C e-widmagot, FCT-tell. s.o\TA=3'/3C "Anwe ngot gigabe nbem-zhewenma, " although one boy 1 . in . the . process . of-pity. s . o . \TA. to e-kedot o mdemoze. FCT-say\AI=3C that. AN old. woman 25 Bama zhe na gete e-wabmat, "Nesh later EMPH EMPH for. sure FCT-see . s . o\TA= 3/3' C contrarily je gyetnam nakwtem mine nta-mnekwe but for. sure talk . back\AI2=3I and like . doing-drink\AI=3I Sure enough, soon he saw him, "He sure talks back, and he drinks a lot, that bad young fellow , " said the old woman. o wshkenigesh, " e-kedot o mdemoze. that. AN young. man=PEJ FCT-say\AI=3C that. AN old. woman 27 Bama zhe na gete shkech later EMPH EMPH for. sure a. while e-wi-ne-iribawat e-bye-bapashkwet FCT-FUT-start . to-sleep\AI=35C FCT-come-whoop\AI=3C weye, "Mbe! O ye o beydwewegze, " someone jeez that ' s . the . one . (AN) CC . sound. comes\AI . 3P e-kedot o mdemoze. FCT-say\AI=3C that. AN old. woman Sure enough, later on when they were going to sleep, they heard someone whooping. "That's him all right, coming yelling, " said the old woman. to to 23 Bama zhe na shkech later EMPH EMPH a. while e-bye-bidge-go jek, "Noko gyetnam FCT-come-enter\AI grandma for. sure ndenniw," e-kedot 1-be.a.man FCT-say\AI=3C After a while, he came tumbling in, "I'm the man, Grandma" said the boy, and he kicked around some pails. gigabe, e-bye-ddegshewat kekoyen. Boy FCT-come-kick. around. s . o. \TA=3/3'C pail=OBl^ 29 Iw se gigabe e-gi-bzegwidzet that.INAN EMPH boy FCT-PST-get . up\AI=3C e-gi-zagjewebnat . FCT-PST-throw. s . o . out\TA=3/3' C So the Crane-boy got up and threw him outside. 30 "Noko, ni VTpi je ga-danet grandma where CC . PST-live . in . a . certain . place\AI=OBI/'=3P ode bye- je-nshiwzet zhode this come-f or . to-be . efficient\AI=3C here "Grandma, where does he live, this fellow who 's come to run our place ? " said the bad boy. edaygo , " e-kedot o live. in. a. certain. place\AI=15 . C FCT-say\AI=3C that. AN to shkenigesh . young . man=PE J 3 J Iw zhe na, "Wi-majin, noses," that.INAN EMPH EMPH FUT-take . s . o . away\TA my . grandchild e-nat o mdemoze, "ekwiyen FCT-say. to. s . o. \TA=3/3'C that. AN old. woman that's. right bba-nshonadzen . around-be . naughty\AI=2IMP "You go away, now, grandchild" said the old woman, "that's right, you go on, behaving any old way. " 32 Iw je ge na e-gi-gizhgennen and also EMPH FCT-PST-raise . s . o . \TA=1/2C e-gish-yanwe ' nan e-wi-bzedwin, FCT- finish-believe. s.o. \TA=1/2C FCT-FUT-listen . to . s.o. \TA=2/1C "/ already raised you, and I can't make you listen anymore. Go on and take care of yourself, now. " wme-bba-bmendezon zhe na zhye. go. and-around-take. care. of . o. s . \AI=2IMP EMPH EMPH EMPH 33 Gde-mgegno zhe na, 2-be.big\AI . I EMPH EMPH e-wi-bmendezyen, " FCT-FUT-take.care.of .o. s. \AI=2C "You're big enough to take care of yourself, " the old woman said to her grandson. to e-nat ossesen mdemoze. FCT-say. to. s . o. \TA=3/3'C grandchild=OBI^ old. woman 34 "Wakokiwek jo as . long . as . the . world. stands not gwi-nme-gkenmasik gmezodanek, 2 -getting. to. be- know. s.o. \TA=DIR=NEG=2/35I 2-parent=PL wzam e-naktemen. too. much FCT-talk.back\AI=2C 35. I ye i that ' s . the . one . ( INAN) wa-wje-zhewebzin, CC . FUT-why-have . happen . to . one . a . certain . way\AI=2P mshikes gge-nme-go turtle 2 . FUT-getting. to.be=say. to. so. \TA=INV. I "You talk back too much, so you will never know your parents. That's why this will happen to you: you will always be called turtle. You'll never have any smarts, because you talked back too much, " he was told. wakokiwek jo as . long . as . the . world. stands not gge-nme-winedbisi , 2 -FUT-getting. to . be- FUT-have . brains \AI . I e-bbich-naktemen, " e-gi-nayek. FCT-so.much-talk.back\AI .2C FCT-PST-say . to . s.o. \TA=PASS=3C 35 Iw je ye i we j-bwa-gkenmat nmezodan and PRED that . INAN CC . why-NEG- know. s . o . \TA= 3/3' C 1-parent o mshike . that. AN turtle That's why the turtle doesn't know his parents. 37 Iw se gigabe e-gi-me-mik jewit that.INAN EMPH boy FCT-PST-continue-work\AI=3C jejakos e-gi-bmenmat crane=DIM FCT-PST-take . care . of \TA= 3/3' C So the little crane kept working, and he took care of his grandmother all winter. gbe-bbon okmesen. through . all . of-winter grandmother-OBl^ to 3g Iw je neko kezhyep o mdemoze and used. to early that. AN old. woman e-widmawat eni FCT-tell. s.o\TA=35/3'C this . over . there . OBV So usually in the morning, the old woman would tell him what she would be cooking. wa-me-zekwet . CC. FUT-continue-cook\AI=3P 39 "Mbish naden, noses," water fetch . s . t . \TI=2/0IMP my . grandchild e-got gigabe. FCT-say. to. s.o. \TA=3'/3C boy "Fetch water, grandchild, " she would say to the boy. 4Q "Gask-wiyas mine mdamnesek dried-meat and corn=DIM=PL nwi-gizzwak e-wi-wisneygo, " e-kedot. 1. FUT-cook. s.o. \TA=1/3I FCT-FUT-eat\AI=15 . C FCT-say\AI=3C "I'll cook dried meat and corn for us to eat, " she said. 41. to On 42. 43. Pene "Wech always why i kedot?" gigabe wegni je that.INAN say\AI=3C boy what yedek, "Pene wech-widmewat must.be always CC . why-tell . s . o\TA= 3/3' C wa-ne-zekwet? " CC. FUT-start. to-cook\AI=3P I me je wi zhe pene that.INAN EMPH but EMPH EMPH always e-kenongot o gigabe. FCT-talk. to. s.o. \TA=3'/3C that. AN boy I je e-byat neko and FCT-come\AI=3C used. to e-nnatagot ni okmesen "Weye FCT-ask. s . o. \TA=3'/3C that . OBV grandmother=OB\^ someone ne ggi-wabma noses," Q 2-PST-see. s . o\TA=DIR. I my . grandchild e-got ni mdemozeyen. FCT-say. to. s.o. \TA=3'/3C that . OBl^ old. woman=OB\^ The boy would always wonder, "Why does she do that? Why does she always tell me what she'll cook?" So it must be that he would always talk to the boy. So when he came back, his grandmother would ask him, "Did you see someone, grandchild? " 44 "Jo," e-kedot gigabe. no FCT-say\AI=3C boy "No, " said the boy. 45 I j© ngodek e-zhede'at gigabe, and one=LOC FCT-think\AI=3C boy "Nda-gi-gkendan wegni yedek 1-MOD-PST-know. s. t\TI=OBJ=OBJ what must.be So once the boy thought, "I must know why he asked me what it is we 're going to eat. " wech-nnatewat i wegwendek CC.why- that.INAN whatever wa-mijyak," e-zhede'at. CC. FUT-eat. s.t.\TI=15P FCT-think\AI=3C to <1 4g Iw je ngodek zhe na mine and one=LOC EMPH EMPH and e-nadet kezhyep i FCT-fetch. s. t. \TI=3/0 (5) C early that.INAN nibish, "Wegni wa-mijyek, jejakos," water what CC . FUT-eat . s . t\TI=25C crane=DIM So in the morning, once again, he went to fetch water and someone said to him, "What are you going to eat, little crane?" e-got weyeyen. FCT-say . to . s . o . \TA=3'/3C someone=OB\/^ 47 "Gete se gwi ' dem for. sure EMPH 2 . bother . s . o . \TA=DIR . I pene e-nnatoyen wegni wa-mijyak. always FCT-ask . f or . s . t . \TI=2C what CC . FUT-eat . s . t . \TI=15P "You sure bother me, always asking what we're going to eat. " 48 Iw je ge je widmownen. Wabgoneshen, and also but tell . s . o\TA=INV=l/2I squash=PE J=PL mine gokosh wi-dgoze zhiw, wabgonesgik , " and pig FUT-mix . things\AI=3 . I there squash=DIM=LOC e-kedot gigabe . FCT-say\AI=3C boy Well, I'll tell you: Squash with a little pork mixed in, " said the boy. to 00 49 Iw se e-wi-gkendek zhye gigabe. that.INAN EMPH FCT-FUT-know. s . t\TI=3/0C EMPH boy 5Q Gego bama zhe na gete . something soon EMPH EMPH for. sure 5 J E-nme-gisekwet mdemoze, FCT- get ting .to. be- finish. cooking\AI=3C old. woman e-bye-bidgeshkak gche-emkwan. FCT-come-enter\II=OC big-spoon So now the boy wanted to know [what would happen]. Something happened, sure enough. When the old lady is almost through cooking, in comes a big spoon. 52 Iw se mdemoze, "Yaa, noses! that.INAN EMPH old. woman EXCL my . grandchild Gi-yajmo negne * 2 . PST-say\AI . I so Iw se e-wi-bkedeygo, iw that.INAN EMPH FCT-FUT-be . hungry\AI=15 . C that.INAN se ye i ga-wje-dne 'monan EMPH PRED that.INAN CC . PST-why-tell . in . a . certain . place=l/2P "Yaa! Grandchild! " she exclaimed, "You told after all! Now we'll be hungry. That's why I told you not to tell what we would be eating, " she said to the boy. e-wi-bwa-yajmoyen wegni FCT-FUT-NEG-tell . about\AI=2C what wa-mi jyego, e-nayet gigabe . CC. FUT-eat. s. t\TI=15.C FCT-say . to . s . o . \TA=PASS=3C boy 53 Iw gshe zhe gete e-gi-bkedewat that.INAN EMPH EMPH for. sure FCT-PST-be . hungry\AI=35C gbe-gizhek, iw gshe e-gi-gkendek through. all. of-day that.INAN EMPH FCT-PST-know. s . t\TI=3/0C e-zhewebek e-bkedek FCT-happen. a. certain . way\ I I=OC FCT-be . hungry\AI=3C Sure enough, they were hungry all day, and so the boy knew what it was like to be hungry. gigabe . boy 54 E-bkonyak e-widmagot ni FCT-be.night\II=OC FCT-tell . s . o\TA=0/3C that . OBV okmesen, "Gego mine yajmoken grandmother=OBl^ don't and tell . about . s . o\TA=0BJ=2IMP When it was night, his grandmother told him "Don't say anything else about what we will eat " wegni wa-ini jyego, " what CC. FUT-eat. s. t\TI=15.C e-nayek . FCT-say. to. s.o. \TA=PASS=3C o o 55 Iw se gigabe e-nagdewendek that.INAN EMPH boy FCT-keep . s . t . in .mind\TI=3/0C ga-zhewebzet . CC . PST-have . happen . to . one . a . certain . way\AI=3C 5g "Nehaw, noses," e-nayek mine kezhyep, okay my . grandchild FCT-say . to . s . o . \TA=PASS=3C again morning " I ngom gego wi ngom yajmoken that.INAN today don't EMPH today tell . about\AI=2IMP So the boy kept in mind what happened. "Okay, grandchild, " he was told again in the morning, "Today, don't tell him what we 're going to eat: I'm going to make corn soup. " wa-mijyak: mdamnabo nwi-gizsan." CC. FUT-eat. s.t. \TI=12C corn. soup 1-FUT-cook . s . t . \TI-OBJ-OBJ. I 57 "Wewene bshe nwi-nagdewendan ngom carefully EMPH 1-FUT-keep . s . t . in .mind\TI . I today neko," e-kedot o gigabe. used. to FCT-say\AI=3C that. AN boy "I'll certainly keep that in mind today, " said the boy. 58 E-nme-zag' ek FCT-in . the . process . of -go . outside\AI2=3C e-kenongot mine weye, "Wegni FCT-talk. to. s . o. \TA=3'/3C again someone what When he went out, again someone spoke to him, "What are you going to eat. Utile crane?" wa-mijyek jejakos?" CC. FUT-eat. s. t\TI=25P crane=DIM 59 "Ni je bshe wa-dodwayek , " what EMPH CC . FUT-do . something . to . s . t . \TI e-kedot je jakos=giga}3e . FCT-say\AI=3C crane=DIM-boy "What are you going to do? " said the Crane-boy. o gQ Babek okmesen e-nnatagot, sure. enough grandmother=OBl^ FCT-ask . s . o . \TA=3' /3C "Ni je na, noses, gi-yajmo ne . " what EMPH my . grandchild PST-tell . about\AI . 31 Q Sure enough, his grandmother asked him, "Well, grandchild, did you tell?" Q\ "Jo noko nda-ya jmosi , no grandma 1-MOD-tell . about\AI -NEG ngi-bkede je ge na we j-bwa-yajmoyen, " l-PST-be.hungry\AI but also EMPH CC. "No, grandma, I didn't tell I was hungry, so that's why I didn't tell, " said the boy. e-kedot gigabe . FCT-say\AI=3C boy 52 Gigosen ge e-gche-mamikmewat fish=OBV also FCT-really-gather . s . o . \TA- 35/3' C e-gwa je ' gonet FCT-be. washed. ashore\AI=OBV"=3C They gathered fish that the waves had washed ashore, and the old woman would roast them. e-mwawat, neko e-bwenet FCT-eat. s.o. \TA=35/3'C used. to FCT-roast . s . t . \AIO=OBV"=3C o mdemoze. that. AN old. woman O to 53 Iw se o gigabe that.INAN EMPH that. AN boy e-gche-mneset pene gigosen pene FCT-really-cut . wood\AI=3C always fish=OBl^ always e-bwenewat . FCT-roast . s . t . \AIO=OBV"=35C The boy gathered lots of wood, because they would always roast fish. 54 Iw gshe zhe neko e-ne-iribawat that.INAN EMPH EMPH used. to FCT-start . to-sleep\AI=35C pene e-gkyekmegot ni okmesen. always FCT- that . OBl^ grandmother -OBV So usually when they were going to sleep, his grandmother would teach him. 55 Ngodek zhe na e-zhgezhgeshek o gigabe e-bkonyak one=LOC EMPH EMPH that. AN boy FCT-be . night\II=OC gechwa zhe na e-nodwat bmekchakoyen , like EMPH EMPH FCT-hear . s . t\TI=35C frog=OBl^ what Once, when he was lying down at night, the boy heard something that sounded like frogs. " What 's happening. Grandma?" said the boy. "Ni je ezhwebek, what CC . happen . a . certain . way\II=OC noko?" e-nat gigalse . grandma FCT-say . to . s . o . \TA=3/3' C boy O gg "Ni je ezhwebzin, noses?" what CC . be . in . a . certain . state\AI=2C my . grandchild e-kedot mdemoze. FCT-say\AI=3C old. woman 57 " Gechwa gshe nin bmekchako nnodwa , " like? EMPH I. EMPH frog hear.s.o\TA -DIR e-kedot gigabe . FCT-say\AI=3C boy "What's the matter, grandchild?" said the old woman. "Seems like I hear frogs. " said the boy. gg "O noses, iw gshe nina zhye beshoch oh my. grandchild that.INAN EMPH EMPH near e-wi-nokmek, " e-kedot mdemoze. FCT-FUT-be. springtime\II=OC FCT-say\AI=3C old. woman "Oh, grandchild, now soon it will be Spring. " said the old woman. 59 "Ezhi pene gwi-nap wech-nawkwek over. there always 2-FUT-look\AI . I CC . towards-be . noon\II=OC gishpen bye-gdegdegankok, if come-be . spotted. clouds\II=OC "Always look there, towards noon. If spotted clouds come, you're parents will be leaving to come back" said the old woman. iw ye i zhye e-bye-majiwat gmezodanek that.INAN PRED that . INAN EMPH FCT-come-leave\AI=35C 2-parent=PL e-wi-byawat, " e-kedot mdemoze. FCT-FUT-come\AI=35C FCT-say\AI=3C old. woman o 7Q Gyetnam zhe e-mnewendek gigabe . for. sure EMPH FCT-be glad\AI2=3C boy ^\ Babek kezhyep e-dokit. sure. enough morning FCT-wake . up\AI=3C 72 'L'h& gshe e-dnabet there EMPH FCT-look . towards . a . certain . place=3C wech-nawkwek gawa zhe na CC. towards-be. noon\II=OC scarcely EMPH EMPH bme-nangodek ode i e-wabshkankok . along-once . in . a . while this that.INAN FCT-be . white . clouds=OC The boy was very glad. Sure enough, in the morning he woke up. He kept looking towards noon, just barely, every once in a while, there would be white clouds. o Y3 Gigabe e-bidgesat okmesen boy FCT-fly. in=3C grandmother -OBV e-widmawat, "Bme-nangodek FCT-tell . s . o\TA=35/3'C along-once. in. a. while ode e-gdegankok, noko," e-kedot thisFCT-be. spotted. clouds\II=OC grandma FCT-say\AI=3C gigabe . boy 74 "I zhye e-bye-majiwat gmezodanek, " that.INAN EMPH FCT-come-leave\AI=35C 2-parent=PL e-kedot mdemoze. FCT-say\AI=3C old. woman "75 Iw se e-dapnat wdekkon that.INAN EMPH FCT-pick . s . t . up\TI=3/0C 3-pail=OBl^ o gigabe; "Igwamsen, noses, gego that. AN boy careful my . grandchild don't wi -y a j moken wa -mi j yego , FUT- tell. about \AI=2 IMP CC . FUT-eat. s.t\TI=15.C nwi-gizsanen bokshegneyek , 1-FUT-cook . s . t . \TI=OAM=OBJ=PL . OBJ yuccapan=PL mko weyas , nwi-bbawe," e-kedot bear meat 1-FUT-mix . things . around FCT-say\AI=3C The boy came flying in and told his grandmother, "Every once in a while, I see spotted clouds, Grandma! " "Your parents are starting to leave, " said the old woman. So the boy took his pail "Be carefiil Grandchild; don't tell what we 're going to eat: I'm going to cook yuccapans with bear meat mixed in, " said the old woman. mdemoze . old. woman 7g Wete se e-mnotwat gigabe really EMPH FCT-like . hearing . s . o . \TA=3/3' C boy That really sounded good to him. okmesen wa-ne-zekwenet . grandmother-OBl^ CC . FUT-start . to-cook\AI=OBV"=3P yy "Jo gshe nde-yajmosi, " e-zhede'at gigabe. not EMPH l-tell.about\AI=NEG. 31 FCT-think\AI=3C boy 7g Gete ga-gish-gwap ' ek i nibish, for. sure CC . PST-f inish-scoop . s . t . up\TI=3/0C that.INAN water e-nnatagot ni nenwen, FCT-ask. s.o. \TA=3'/3C that . OBl^ man=OBl^ "For sure, I won't tell " thought the boy. After he dipped into the water, that man asked him "What are you going to eat, little crane?" o "Wegni wa-mijyek jejakos," what CC. FUT-eat. s. t\TI=25P crane=DIM e-nayek gigabe. FCT-say. to. s.o. \TA=PASS=3C boy 79 Babek okmesen e-nnatagot, sure. enough grandmother=OBl^ FCT-ask . s . o . \TA=3' /3C "Ni je na noses, gi-yajmo ne?" what EMPH my . grandchild PST-tell . about\AI Q e-got . FCT-say. to. s.o. \TA=3'/3C Sure enough, his grandmother asked him, "Well, grandson, did you tell?" gQ "Jo wi zhe na, noko," e-kedot gigabe. no EMPH EMPH EMPH grandma FCT-say\AI=3C boy "No, I did not. Grandma, " said the boy. Iw se na e-gi-mno-wisnewat, ngodek zhe na that.INAN EMPH EMPH FCT-PST-good-eat\AI=35C one=LOC EMPH EMPH o gigabe e-mlkwendek mine that. AN boy FCT-think . of . s . t . \TI again So they had a good meal, and the boy once again thought of what his grandmother had said. ga-kednet ni okmesen . CC. PST-say\AI=3 ' P that . OB\^ grandmother-OBt^ g2 "Gete shke na wi-bye-wdegankot, " for. sure that's. it EMPH FUT-come-be . a . spotted. cloud=OI e-zhede'at gigabe, "kezhyep nwi-doki FCT-think\AI=3C boy early 1-FUT-wake . up\AI "That's right, the spotted clouds will come, " thought the boy. "I'll wake up early and look for the spotted clouds. " o 83. e-wi-nde-wabdemen e-wi-gdegankok . FCT-FUT-1-see. s . t\TI FCT-FUT-be. spotted. clouds=OC " Iw se zhye yedek e-wi-byawat that.INAN EMPH EMPH must.be FCT-FUT-come\AI=35C nmezodanek , " e-zhede ' at . l-parent=PL FCT-think\AI=3C "So now must be my parents will come, " he thought. g4 Iw se mine wdekkon e-nanat, that.INAN EMPH again 3-pail=OB\^ FCT-f etch . s . o . \TA=3/3' C "Noses, mdamnek nwi-gizswak, seksi-wiyas my . grandchild corn=PL 1 . FUT-cook . s . o . \TA=1/3I deer=meat So he took his pail again, and he was told "Grandchild, I'll cook corn and mix in some deer meat. " nwi-bba-bwe . " e-nayek gigabe. 1. FUT-around-roast\AIO. I FCT-say . to . s . o . \TA=PASS=3C boy g5 Jo zhe na wbezdewasin okmesen, not EMPH EMPH 3 . listen . to . s . o . \TA=DIR=NEG. I grandmother-OBI/ e-bbich-nchiwenmot e-wi-wabmat He didn't listen to his grandmother, he was so happy about seeing hisparents. When he came to the FCT- so .much-be . happy\AI=3C FCT-FUT-see . s . o\TA=3/3' C wmezodanen, ibe e-byat zibik babek zhiw 3-parent=OBl^ there FCT-come\AI=3C river=LOC sure. enough there river, sure enough, someone asked him, "What are you going to eat, little crane?" mine weye e-nnatagot, "Wegni again someone FCT-ask . s . o . \TA=3'/3C what wa-mijyek jejakos?" e-nayek. CC. FUT-eat. s. t\TI=25P crane=DIM FCT-say . to . s . o . \TA=PASS=3C gg "Da she ma je se ge gin EXCL? EMPH also 2 . EMPH gdezhwebes mteno 2 . be . in . a . certain . state . I only e-wi-nega'et gich-bmazdi. FCT-FUT-abuse . s . o . \TA=2/3C 2 . fellow-creature "Boy, you certainly have a way of abusing your fellow creatures. gy E-penmen e-nshiwzin FCT-depend.on. s.o. \AI=2C FCT-be. fast\AI=2C e -bme -mnene t FCT-along-take . away. from. s.o. \TA=2/3C You're dependant and swiftly take away what your fellow creatures would eat. wa-mijet gich-bmadzi. CC. FUT-eat. s.t.\TI=3/0P 2. fellow-creature gg Mdamnesek, gask-weyas, wi-dgode corn=DIM=PL dried-meat FUT-be .mixed. in\II . 01 ge-nabjetonen je zhi . " 2?-use. s. t. \TI=2/05. I but there She's going to cook a little corn mixed with dried meat, whatever you may do about it. " 89. iw se gigabe e-gi-myazhewit So the boy once again did wrong. that.INAN EMPH boy FCT-PST-do . wrong\AI=3C and 9Q Iw se bapich e-nnatewat that.INAN EMPH now. and. then FCT-ask . s . o . \TA=3/3' C ni okmesen, that. OBV 3 . grandmother-OBI/^ "De ni e-bbich-zekwenet?" how. far FCT-so . long-cook\AI = OBV"=3C g\ " Iw zhe gaga," e-nayek that.INAN EMPH almost FCT-say . to . s . o . \TA=PASS=3C gigabe, "Shkwademek nwi-jejibdep noko," boy door=LOC l-FUT-sit\AI . I grandma e-kedot gigabe. FCT-say\AI=3C boy Every once in a while, he asked his grandmother, "How soon until it is cooked? " "Almost. " The boy was told. "I'll sit by the door. Grandma, " said the boy. 92 Gche-mtek e-gkegkebet. big-stick FCT-sit . hidden\AI=3C 93 "Ni je ezhwebzin, noses?" what CC . be . in . a . certain . state\AI my . grandchild e-nayek . FCT-say. to. s.o. \TA=PASS=3C He sat hiding with a big stick. "What's the matter, Grandchild? she asked him. 94 "Jo ne ggekendesin, noko? Gzhate not Q 2-know. s . t\TI=NEG=OBJ grandma be . hot . weather\II ma zhe na ode , noko . Gechwa zhe na gde-gkendan . but EMPH EMPH this grandma like EMPH EMPH 2-know. s . t\TI "Don't you know, Grandma? It's hot here. Grandma. Seems like you would know." 95 Iw zhye e-mnokmek ggi-ket, ge na noko. that.INAN EMPH FCT-be . spring\II . I 2 . PST-say\AI . I also EMPH grandma "It's getting to be spring, you said. Grandma. " 95 I je ye i wech-gche-gzhatek, " that's it CC . why-really-be . hot . weather\II=OC e-kedot gigabe . FCT-say\AI=3C boy 97 Iw zhe na zhye e-bme-gwashmat that.INAN EMPH EMPH EMPH FCT-along-take . s . o . of f . f ire\TA=3/3' C ni wdekkon o mdemoze. that.OBl^ 3-kettle=OBl^ that. AN old. woman "That's why it's getting to be hot, " said the boy. So then, the old lady took up her kettle. 9g Bama zhe na e-bye-bidgeshkannek soon EMPH EMPH FCT-come-enter . with . body\AI = OBV"=0C That big spoon came reaching in. iw gche-emkwan. that.INAN big-spoon 99 Gigabe babek boy sure. enough wmetgom ga-nwedsat 3-stick=P0SS CC. PST- take. hold. of . s.o. \TA?=3/3'C e-gi-baskeknadek i gche-emkwan. FCT-PST-split. s. t. \TI=3/0C that.INAN big-spoon So the boy grabbed the stick and split that big spoon. 100. Iw se e-gi-zagjebozot gigabe. that.INAN EMPH FCT-PST-run . out\AI=3C boy So then the boy ran out. 101. 102. Ibe zhe na ga-wje-byat there EMPH EMPH CC . PST-where-come\AI=3P e-zhe-gche-majit e-byat ibe FCT-in. a. certain. way-really-leave\AI=3C FCT-come\AI=3C there jajibdebet e-ne-wabet bzhe ibe sit\AI=3C FCT-start. to-see\AI=3C EMPH there e- je-gdegankodnek . FCT-there-be . spotted. clouds\II=OBV"=C Pene zhe na gego e-nshet, always EMPH EMPH something FCT-hear . in . a . certain . way\AI=3C wika zhe na e-bzegwit finally EMPH EMPH FCT-stand. up\AI=3C we na pi e-zhewebzet so. far FCT-CC.be. in. a. certain. state\AI=3C ibe e-nesmegagwet there FCT-f ace .in. a. certain. direction\AI=3C wete zhe na e-bye-mkedewangok . for. sure EMPH EMPH FCT-come-be . black . clouds\II=OC He ran to the place where he had come from, and when he arrived, he sat down and looked, and there were spotted clouds! He began to hear something, and finally he stood up and faced that way— for sure, a black cloud was coming. 103. "Wzam ne wi zhe na gi jejakok too. much Q EMPH EMPH EMPH those. AN crane=PL a-yawik," e-kedot gigabe . MOD-be. a. certain. thing\AI=35I FCT-say\AI=3C boy "Is it too much? It must be the cranes! " said the boy. 104. Bama zhe na gete jejakok e-byawat soon EMPH EMPH for. sure crane=PL FCT-come\AI=35C i se mine e-gche-gwagwaskze ' ot that.INAN EMPH again FCT-really- jump . up . and. down\AI=3C Soon, sure enough, the cranes came and again the Uttle crane jumped up and down, he was so happy. e-bbich-nchiwenmot o jejakos. FCT-so. much-be. happy\AI=3C that. AN crane=DIM 105. Bama zhe na wgyeyen mine soon EMPH EMPH 3.mother=0BV and osen e-bme-nde-wabmagot, 3 . f ather=OBt^ FCT-along-try-see . s . o . \TA=3/3' C "A, ngwese, ah 1 . dear . son Soon his mother and father were looking for him. "Ah, my son, we are very glad you are still alive, " said his parents. to i se gweyen e-bmadzeyen megwa," that.INAN EMPH good FCT-live\AI=2C still e-got ni wmezodanen. FCT-say. to. s.o. \TA=3'/3C that . OBl^ 3 . parent=OBl^ -ir\f: "Neshpa se na nokmes jo wi zhe o-- if. not. for but EMPH 1 . grandmother not EMPH EMPH that. AN ndo j-bmadzesi o wabozo 1. reason-live\AI=NEG. I that. AN rabbit "If it weren't for Grandma, I wouldn't be— I couldn't live on just that rabbit you abandoned me with, " said the boy. ga-ngedmayek , " CC. PST-leave. s.t. behind. for. s.o. \TA=PASS=3P e-kedot gigabe . FCT-say\AI=3C boy 107 "'"*' ^^^® ^® ''^ ° e-kwadsokazot So that is as far as the story goes, ■ that.INAN EMPH also EMPH that. AN FCT-be . so . long . as . a . story\II = OC about Uttk Cram misbehaving. o jejakos e-gi-gyebadzet . that. AN crane=DIM FCT-PST-misbehave\AI=3C 108 ^"^ ^^^® ^^°- That's all. that.INAN EMPH a . certain . amount