THE KAPSIKI LANGUAGE Thesis for the Degree of Ph. 'D. MICHIGAN STATE UNIVERSITY DAVID MARTIN SMITH 1969 This is to certify that the thesis entitled THE KAPSIKI LANGUAGE presented by DAVID MARTIN SMITH has been accepted towards fulfillment of the requirements for M; degree in- flthrOpOIOgy Date Jul" 231 0-169 ‘AR? Miciiig :n Sta tee University ‘5? r; BINDING BY ' L. mm; & SIIIIS' . ‘ nnnK RI‘IBFDY INC. .CZRI" l; OER?) l ‘ I illI - The thes kgage 'ith s; sis of the gran:- atzezpt is made ion of action, The are: stratification: coding mechanis: mi :1; as exp-e: “agate, so vie lag the world 0: 50m symbols . An ethn ‘1 32?; social f N“‘.’“f;n‘ ~"*'«=Q in the v .- " ‘ 5:5 M. of the s ABSTRACT THE KAPSIKI LANGUAGE By David Martin Smith The thesis represents an original analysis of the Kapsiki language with special emphasis on the verb system. Through an analy- sis of the grammatical and semological structure of the verbs, an attempt is made to specify something of the Kapsiki speaker's concep- tion of action, which this structure reflects. The grammatical and semological analysis is based on a stratificational model of language. That is, language is viewed as a coding mechanism, composed of strata, standing between the world of reality as experienced by speakers and the sounds used in speech. Language, so viewed, enables a speaker to encode information concern- ing the world of reality into units of sound and conversely to decode sound symbols. An ethnographic sketch, noting the salient ecological, cultur- al and social features of the Kapsiki's total context of action is provided in the first section. This is followed by a rather detailed sketch of the structure of the language in general, including a de- scription of word and construction classes. The third section contains an analysis of the verbs, first, as grammatical realizations of the semology and secondly, as semologi~ cal realizates. Several dimensions of actions, perceived by speakers, and symbolized in the semology are examined. Finally, several suggestions are tendered as to what this analysis of verb semology reveals of the Kapsiki's view of this seg- ment of reality. THE KAPSIKI LANGUAGE by David Martin Smith A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Anthropology 1969 The rese rier provLsiom Daring ‘i in a number literati Brethx Its. H‘mter des :c her language cf the Catholi trier of coca :!5 I V nioently on Q . ~ Ian: zered by v, A“. 4 lb. Maud»: .uu‘eu . P54 ... Mon: ~c : \v .5?" UDC‘ Chayfie ‘ ACKNOWLEDGEMENTS The research on which this dissertation is based was conducted 'under provisions of a Fulbright Hays research grant, number FH6-lh9. During our stay in Cameroun the following rendered invaluable aid in a number of ways. The Reverend and Mrs. Lyle Hunter of the Lutheran Brethren Mission provided facilities, help and friendship. .Mrs. Hunter deserves special thanks for permitting me free access to her language notes. In addition the Reverend Father Andre Durier of the Catholic mission at Siir proffered friendship and help on a number of occasions. It will be gratifying if this analysis should incidently prove helpful to these people in their struggles with the language. I am also indebted to the many laborious hours of help ren- dered by Mr. Luc Sunu my field assistant as well as the services of JMr..Matthieu Kweji Zama and Mr. Joseph Zera Mpa. It is impossible to mention all of the Camerounian friends who helped us. Many thanks are due my Ph.D. committee for their patient work and encouragement, to Professor Alfred B. Hudson my chairman, to Pro- fessor Charles C. Hughes, director of the African Studies Center and to Professor Charles Kraft. Finally, I owe a real debt of gratitude to my wife, Janet F. Smith, for her help in typing manuscripts. ii “If A wk “'3 OF SY‘ 'v in \L . NH C! can Uafi ‘ TM 2‘ 1\ 2‘ —‘\ 1\ II‘ 0 I O O WkLll~I§Wil IL -14. xi TABLE OF CONTENTS ILIST OF FIGURES LIST OF SYMBOLS AND ABBREVIATIONS SECTION ONE - THE PEOPLE . . . . . 1.1 GENERAL BACKGROUND . . . 1.1.1 Introduction . 1.1.2 Where They Live . . . 1.1.5 Where They Came From . . 1.1.h Who They are Related to 1.2 HOW A KAPSIKI SEES HIMSELF l .MARGI or Other'. . . . .2 Margi or Fulbe . . . . . 3 Pagan or Christian . . . 1.3 How HE ORDERS HIS LIFE The Village . . . . The Clan . . . . . . . The Family . . . . 1.h E Ceremonies Marriage Summary . . l 2 .5 . .h Blacksmiths and Slaves 5 6 7 SETTLES HIS DISPUTES . 1.h.1 Kinds of Disputes . . 1 4.2 Structures and Processes 1.5 HE RELATES TO REALITY . . . 1.5.1 Religious Beliefs . 1.5.2 Birth and Death . . . . 1.5.3 The Smith . . . . . . . 1.6 HE MAKES A LIVING . . . . . . 1.6.1 Annual Cycle . . . . . . 1.6.2 Division of Labor . . . iii PAGE vi vii }_J CD NO\O\ F— FUJNI-J ll 12 1h 16 18 18 18 21 2h 2% 17 so 30 SO 52 1.1 as Io; 2o rm 2.1 952:: Oil N0 - U6“ CL \\ WP? |. _. .o.-o4 . o zulwzyzu .«I‘ a . l. ....p..... fig ~LJ SECTION TWO- THE LANGUAGE . . . . . . . . . . . . . . . . . . 37 2.0 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 37 2 .l PHONOLOGY . . . . . . . . . . . . . . . . . . . . . . ’40 2.1.1 Vowels . . . . . . . . . . . . . . . . . . . . . NO .1.2 Consonants . . . . .'. . . . . . . . . . . . . . Al .1.3 Morphophonemics . . . . . . . . . . . . . . . . . A7 .1.h Tone and Stress . . . . . . . . . . . . . . . . . A9 2.2 WORD CLASSES . . . . . . . . . . . . . . . . . . . . . 54 0 Introduction . . . . . . . . . . . . . . . . . . 5h 1 Nominals . . . . . . . . . . . . . . . . . . . . 55 2 Verbs . . . . . . . . . . . . . . . . . . . . . 63 .3 Qualifiers . . . . . . . . . . . . . . . . . . . . 7O 14. 5 NNN Relators . . . . . . . . . . . . . . . . . . . . 76 Expletives . . . . . . . . . . . . . . . . . . . 83 NNNNNN 2.3 CONSTRUCTION CLASSES . . . . . . . . . . . . . . . . . 88 .3. l Phrases . . . . . . . . . . . . . . . . . . . . 88 .3 2 Classes . . . . . . . . . . . . . . . . . . . . 93 .3. 3 Sentences . . . . . . . . . . . . . . . . . . . . 100 .3 .A Discourses . . . . . . . . . . . . . . . . . . . 103 SECTION THREE - VERBS . . . . . . . . . . . . . . . . . . . . 108 3.1 GRAMMATICAL FEATURES . . . . . . . . . . . . . . . . . 110 3.1.1 Verb Roots . . . . . . . . . . . . . . . . . . . . 110 3 1.2 Verb Stems . . . . . . . . . . . . . . . . . . 112 3.1.3 .Markers and.Affixes . . . . . . . . . . . . 120 3.1 3 .h Uninflected and Unmarked Stems . . . . . . . . . 128 .1. 5 Auxiliary Verbs . . . . . . . . . . . . . . . . . 129 3.2 SEMOLOGICAL REALIZATIONS . . . . . . . . . . . . . . . 132 3.2.1 Temporality . . . . . . . . . . . . . . . . . . . 13h 3.2.2 Aspect . . . . . . . . . . . . . . . . . . . . . 136 3.2.3 Modality . . . . . . . . . . . . . . . . . . . . . 1AO 3.2.h Roles . . . . . . . . . . . . . . . . . . . . . . 1&5 3.3 A SEMDLOGICAL NETWORK . . . . . . . . . . . . . . . . 150 3.3.1 The Network . . . . . . . . . . . . . . . . . . . 150 3.3.2 The Discourse . . . . . . . . . . . . . . . . . 155 3.3.3 Glossary of Notations . . . . . . . . . . . . . . 157 3 3 .h Realization Rules . . . . . . . . . . . . . . . . 158 iv SECTION FOUR flaw": . HELIMM r13 Ito's-m 5...“.‘1'3 . 5.A SEMOLOGICAL REALIZATES . . . 3.h.l A Dynamic View of Reality . 3.h.2 Spatial Dimensions of Action . 3.h.3 Temporality and Spatiality SECTION FOUR — CONCLUSION BIBLIOGRAPHY .APPENDIX - THE CORPUS PAGES 160 162 162 165 166 169 171 LIST OF FIGURES Figure 1, Kapsiki Social Organization. Page 5 Figure 2, Consonant Phonemes. Page A3 vi u n r' ”be: trenatlons a I I en: / / en: I I en: I II er, / in: f/ o I it em I“ ‘ see A S; 2’6: tr «5-,, bvqu CC h‘. 4-H Ve 4 «ea: , (16 “A 5J0: CC SYMBOLS AND ABBREVIATIONS Except where specifically noted in the text the following ab- breviations and notational conventions have been used in this thesis. pro spc prep conj dem C 001‘ /V/ /c/ /¢/ enclose phonetic material. enclose phonemic units. enclose morphemic units. enclose morphophonemic units. mark semological units. indicates the following is a morphological realization. indicates the following is a lexemic realization. indicates that the following is a semological realization. is realized as. In the text enclose parenthetical material, in examples in- dicate opitional units. verb, with subscripts, e.g., V2 indicates class of verbs. noun, with subscripts, e.g., N2 indicates noun class. pronoun. specifier preposition conjunction verb marker demonstrative coordinator vowel consonant syllabic consonant marks beginning or ending of an utterance. (underlining) sets off Kapsiki units in the English text. vii enclc enclc I d orcer The I See also the : self 2 E U ' i 1 ‘ t) \n n) (n I (m (D \L) ( N U) ru w on t D '— \fl —4 I " " enclose English glosses in the text. enclose English explanations or tags in the text, for example, 22’ 'Consecutive Sequential Indicator'. 25233_ The ordering of examples is as follows. If the example is not too long the Kapsiki appears on the left, an idioma- tic English gloss to the right, the text reference in parentheses to the extreme right with a literal English gloss directly below the Kapsiki. If the example is long the idiomatic English gloss is found below the literal one 0 See also the following explanations of abbreviations in the text it- self: Page 52 symbols used in the tone examples. Page 93 special conventions used in Construction Class examples. Page 157 glossary of notations used in the Semological Network. viii 1.. RH...‘ I? . I'. ~:.‘E—‘ . .‘ U ‘ 8‘ 1.1.1 Introd \ The Ki rte-1:28 hill 1 T's-ether or no‘ .9 '3: ' ‘ ;. xgerla 82C 391‘ part Of t 2:559 wages 11:72 that of t :3".- ' I ..:..,-brot:1er ' : “ ' . 3' . a lIlSCRRCGI .ccuses on the es: only vi“ mi This tr Iagpo ‘ .301, of Hog as of My » 0 JE‘. “ :5. cent of 3‘ fi m" . _, A ~ e..Q at Moire H‘ i..e ’ ” “an” Pa::‘:& ‘ 0 «Li. rfithI‘ a . | m- ‘\*1entlng 5".8 SECTION ONE THE PEOPLE 1.1 GENERAL BACKGROUND 1.1.1 Introduction. The Kapsiki are a tribe of between 22 and 25,000 who inhabit the Mandara hill region on the Cameroun side of the Nigeria-Cameroun border. ‘Whether or not the several thousand people living in the adjacent area of Nigeria and speaking the several dialects of Higi, should be consid- ered part of the same tribe is a question still not settled. Some of these villages use a dialect that is evidently mutually comprehensible with that of the Kapsiki and most of the villages belong to the same enemy-brother village complex as those in Cameroun. However, except for a few instances in which comparative data are introduced, this paper focuses on the people living in Cameroun and the term Kapsiki will be used only with reference to them. This tribe is the main inhabitant of the administrative division (canton) of Mogode which is ruled by the Lambdo echo resides in the vil- lage of Mogode. The canton is a sub-division of the Margui-Wandala Department of Eastern (ex-French) Cameroun and the department capitol is located at Mokolo. The name Kapsiki, rarely used by the people themselves, is a transliteration of the verb psukg which describes the action of Guinea corn fermenting in the process of beer making and the prefix kg; meaning something like, "the people who". Thus this appellation, evidently first used by the Fulbel invaders, means, "those who make beer." Indeed, the preparation, distribution and drinking of beer is an important e1- ement in Kapsike life. lThe term Fulbe is used here for the ethnic group known in the literature as Fula, Fulani, Peul, etc. Fulbe is the term they use in reference to themselves. (cf., 1.2.2 for a fuller discussion.) 1 The pet must then: and, in a nan be simply Kg; it as come i '< or 4.31 is useo I3) tze flaps; E 9 g ' ‘ fl .0..(. finger: ‘ .reo- feet a: ~f. ‘ . Sun‘s ma n3 Tu I}: - «.88, 881:;- 5!- x R“ x“.- O u Oi Set; 73:): \‘Eat Lola. . “S Q: 5‘): M‘tp” «. as ”Se c ."~-.. . are C lii't . ' f‘ ._ WHie E "‘8; ‘ The people ordinarily refer to themselves as Margi, both to dis- tinguish themselves from neighboring tribes (but including the Higi) and, in a narrower sense, in opposition to the Fulbe who may, in fact, be simply Kapsiki turned Moslem. I use the term Kapsiki because; (1) it has come into general use with reference to this group, (2) the term Margi is used in the literature for another large tribe in Nigeria and (3) the Kapsiki themselves are now accepting and using this name. 1.1.2. ‘Where they live. The Canton of Mogode is located on a plateau about A,000 feet in altitude, 12 to 15 miles wide from west to east and at least twice as long. Both the eastern and western edges of the plateau are com- posed of small but fairly rugged mountain ranges which give way to vast p1ains--on the west the plains of Nigeria and on the east the plain of Maroua. The plateau itself is dotted with rather unique "mountains" ‘which are in fact the granite cores of very old extinct volcanoes. These striking bosses rise abruptly, often to a height of several hun- dred feet and those which are located near villages usually have reli- gious and historical significance for the village. The plateau floor is relatively level, free from rocks and trees, being not badly suited to agriculture. However, the villages are strung out in two roughly parallel lines in the hills at either side of the plateau. These were the most easily defended from the mounted Fulbe raiders who invaded the territory in the last part of the 19th century. Even today most planting is done on terraces built on the rocky sides of the hills. The area in between is considered bush and is inhabited only by a few pastoral Fulbe and some Kapsiki who have recently left the sanctuary of the mountains. The vegetation is of the savanah type characteristic to this kind of semi-arid climate. It consists of coarse grasses, several varieties of shiny—leafed shrubs, some thorn trees and a few baobobs. During the dry season the bush and fields are burned off. The dry season starts the first part of October and heavy rains don't come again until the month of June. This long dry period with a consequent dearth of surface water is probably the most persistantly troublesome p.138 ieal with. Give rainfall of made either the leng‘ cos repercussio Water fc units of the c‘ arrest in some non the pools in from resid a «1.91188 ov er . u.‘ of I-Zaosiki l i f .. In ' - ‘ I‘dségs E-e ‘- :- ..rect beari. L :h. .u&. ' er 1=~- t C . i; I‘LE ‘Q‘Q; troublesome physical feature of the environment the Kapsiki have to deal with. Given the short rainy season with a total average annual rainfall of under 25 inches, it is obvious that slight reductions in either the length, regularity or quantity of the rains will have seri- ous repercussions on the food supply and economy of the people. Water for domestic use is especially a problem during the latter months of the dry season. By February the deep wells dug by the gov- ernment in some villages have dried up and the villagers are dependent upon the pools remaining in swamps and stream beds, which are often far from residences. Rituals to ensure adequate water supplies and disputes over access rights to available sources are common features of Kapsiki life. 1.1.3 Where they came from. The ultimate origin of the Kapsiki is not a matter to be treat- ed in this paper but certain histories of recent migrations have direct bearing on the social and linguistic situation presently exist— ing. The Kapsiki live in a number of rather scattered villages with populations of from less than 500 to more than 1,500. These villages are arranged in complexescxfbrother-enemy groups. Brother villages supposedly have a common ancestor while enemy villages have been tra- ditional rivals. To this day membership in one of the two groups affects the treatment of intervillage disputes. Mogode, along with a number of villages in Nigeria are brothers, in opposition to the "enemy" villages of Kamale, Guria, Sedakote, Rufta and Sir. Apart from the fact of their alliances, these groups exhibit other peculiarities. In Mogode, for example, the dialect of the lan- guage used does not have a sex-gender distinction in the third person personal pronouns. Speakers living in the villages mentioned as members of the rival group regularly make such gender distinctions. Apart from this distinction and a number of lexical and tonal differ- ences, the dialects are not very divergent. Hawever, the two groups possess radically differing explana- tions of origin. The residents of Mogode claim that their ancestors came originally from a village nearly 100 miles to the eaSt, near Maroua. The inhabitants of the other group of villages trace their ancestory in the opposite direction, about 100 miles to the west, near Mubi, Nigeria. There is evidence to suggest that in general terms these claims of origin may be accurate and further, to indicate that the Mogode group came later than the others, splitting the first settlement in two parts. The implications for such a convergence of widely separat- ed people who have come now to speak the same language, on an under- standing of language change, although intriguing, are beyond the scope of this study. However, some awareness of the situation will shed light on the linguistic description to follow, which will be based almost exclusively on the Mogode dialect. 1,1,h Who they are related to. Kapsiki is a Chadic language (Greenberg 1966. A6) and thus a member of the Afro-Asiatic language family. It is closely related to Matakam spoken to the North and northeast and to Bana, spoken to the south and southeast. It does show some striking differences to these languages, however, for example, the noun class system so important in Bana is entirely lacking in Kapsiki. Its relationship to Higi on the west has already been mentioned. Both Higi and Kapsiki are closely related to the Bura-Margi languages to the northwest and are probably members of the same sub-group. The material culture and social-organization (eg., the signifi- cance of the blacksmith in the society) show close affinities with the other mountain peoples of north Cameroun. In the literature all of these people are often discussed together under the rubric of Kirdi. 1.2 HOW.A KAPSIKI SEES HIMSELF. With his identity, furnished by this sketch of his physical and ethnic environment, an attempt will follow to explain how the Kapsiki views himself. An understanding of this will help explain why he re- lates to others in the ways that he does. It appears that he sees the people is his universe as members of certain quite rigidly defined classes. These classes are identi- fiable by certain objective referents and form a series of increasingly see ‘er bina.‘ Evenings . lines.) Cro. categories ”:3. 1'"! '6‘” VJ Us. broke iiscussei in CJ 8.!) H Ch 1:: t‘-—lnn.K- Lite; / e/ ( ( ,' ()Inllll)l'l «:9 r' :1 (”ALT zera /s/ /s i?” ck" / 3’1 mm 3“ 4m a”, smaller binary oppositions. These can be conceptualized as horizontal groupings. (In the diagram they are separated by solid vertical lines.) Crosscutting this classification is a series of vertical categories based on residence and lineage affiliations (represented by the broken horizontal lines). The first set of categories will be discussed in this section and the latter in section 1.3. MARGI +Non-MARGI—+ Ie———Margi Fulbe—1‘ f— Pagan ———+— Mission maze l H - tt fig Chlef p__-___._-_---_-__J T.4 gacé e4c> a) ------ - ------ 1 :3 g’ makweji {Dwi L"""'l"‘“""F“""“‘ makwiyi ...... p - o‘cocui zeramba N 5-! 4--------I--\-----I g m malwamté (5)5 A a . J <32; mAva O h- —————— I-- ————— -I kwéganji rége ........ _..__._--. "smiths" ____-___J m6va Jamel. ........ T.-----..--..- ._..-_._-_ xwa mté "'enemy" Figure l Kapsiki Social Organization 1.2.1 MARGI or Other. The first distinction the Kapsiki makes is between the Kapsiki tribe and everyone else. In articulating his identification at this level he would use the term "MARGI" (here capitalized to distinguish it from the same term used at the next level, indicating a contrast 'with Moslem.) He includes not only the residents of Cameroun in the category of MARGI but his brothers in Nigeria, the Higi. MARGI is defined by several components, including (1) a common language-~although the dialect differences may be significant enough to justify postulating more than one language from an objective lin- guistic point of view, (2) a common or related history-~again from his point of view, (3) certain similar but differing in detail, rituals such as marriage and initiation practices, and (u) the fact that the MARGI constitute the primary resevoir of potential marriage partners. The entity described by the referents outlined here is what we have termed the tribe. The Kapsiki has a keen tribal awareness, en- hanced somewhat by the prominance it has been given in recent attempts to attract tourists to the area and the rather condescending attitude expressed toward the Kapsiki by both the present administration at Mokolo and the larger Matakam tribe to the north. 1.2.2 Margi or Fulbe. To the average villager a more significant distinction is be- tween what he terms the persunfi, that is, Fulbe and the Margi or non- Mbslems. As used here Fulbe is synonymous with Moslem. It includes the Fulbe (ethnically) who represent the administration in the villages and the cattle Fulbe who live in the bush and any Kapsiki who has con- verted. In addition to these there is a fairly large Fulbe population in the village of Mogode attached officially or unofficially to the Lambda or canton chief. This population includes, Malams, dancers, musicians, prostitutes, the Lambdo's councilors, chauffeur, bodyguards and their dependants. Many of these live off the largess of the Lamb 0 who extracts tribute from the non-Moslems. For a Kapsiki, becoming a Moslem, involves a drastic change of cultural identity. He must (1) be circumcised, (2) renounce pagan sacrifices, (3) observe taboos on eating pork, (A) observe the fast ofRaoadan and ( traiition. Toe to prayer and (" {5‘} ordinarily 1 As a res work in the fie to cash crops 1 table, he will slight 1y hig‘ne litigation bro 58 Mo: 35733368 to . tr; 3 habitua Larido to dis Notably For: Furth. the Sovernme the select, f hierarchy . 7 of Ramadan and (5) allow his children to be instructed in the Koranic tradition. In addition he is supposed to (6) observe the moslem calls to prayer and C7) refrain from drinking beer. Furthermore, he will (8) ordinarily relocate in the Fulbe part of the village. As a result of his conversion he will no longer be expected to work in the fields and his wives will try to confine their cultivation to cash crops and gardens, he will be able to eat at the Lambdo's table, he will be exempt from public labor (but is supposed to pay a slightly higher head tax) and can expect preferential treatment in any litigation brought by a non-Moslem. The Moslem finds it difficult to translate this "political" advantage to economic. As a result of his disinclination to cultivate and a habitual unwillingness or inability of the administration or Lambdo to dispense the money promised them, the Moslems as a class are probably worse off economically than either the pagans or Christians. Furthermore, despite the stature the Fulbe have in the eyes of the government, this is in effect a closed class. No real attempt is made to convert the pagans in general but real pressure is brought on the select few who have been "tapped" to rise in the governmental hierarchy; 1.2.3. Pagan or Christian. The Margi, or non-Mbslems, constitute by far the bulk of the Kapsiki tribe. This group is divided into Christians (Mission) and Pagan. There are two Christian Missions working among the Kapsiki. An American Protestant group has a small station with an occasional resident Missionary family at Mogode, and the Roman Catholics have a large mission with a staff of 6 or 7 (mostly French), a dispensary, church and school at Sir. Neither missions--for different reasons-- expect or demand a very traumatic conversion experience in terms of culture change. Among the things which differentiate the Christians from the pagans are, (l) a reluctance to participate in sacrifices, (2) among the protestants a de—emphasis on beer drinking and (5) for both prot— estants and catholics a sometime enforced prohibition against taking a second wife. No change in residence patterns food taboos or participation i Whereas tage, convex-sic There are sever tion of vealzh For the Chris: Status synboh {5m mehir Furthermore, 1 Elards, err-an Christians fi AS a 1 “he“. the . large, The ““19 mainly. 8 participation in village rites is expected. Whereas converting to Islam confers certain "political" advan- tage, conversion to Christianity often results in economic advantage. There are several reasons for this. In the first place, any accumula- tion of wealth by the pagans would be used to procure additional wives. For the Christians, on the other hand, it can be used to acquire either status symbols (radios, metal roofing, market goods) or capital goods (sewing machines, bikes-~which are rented, wholesale marketable goods.) Furthermore, opportunities for employment by the missions as houseboys, guards, errand boys and informants are ordinarily offered to the Christians first. As a result of the economic consideration, and a number of others, the Christian population in parts of the Kapsiki area is quite large. The informants for this study (as noted in the appendix) in- clude mainly Christians and Pagans. 1.5 HOW HE ORDERS HIS LIFE The vertical categories intersecting those discussed above, affect all Kapsiki to some degree. That is, in determining how they will order their relationships with others, pagan, Moslem or Chris- tian alike, they are concerned to know village and clan affiliations. Although this is true in principle there are certain inherent restric- tions on the compatibility of the two sets of categories. For example, the significance of clan affiliation is greatly attenuated for the Mbslems and also the place of residence for a Moslem is, in most cases, Mogode. 1.5.1 The Village. There are about 18 Kapsiki villages. The villages are not com- pact but are spread over rather wide areas. They are composed of polyg- ynous family dwellings surrounded by gardens and connected by footpaths often very picturesquely bordered with euphorbia. The villages are always located either in the mountains or at the foot of a mountain. Village affiliation is extremely important to the Kapsiki and is, at the village level, viri-patrilocal. A person belongs to the village of his who marry outs in times of fa tang-rage" anc‘ Each vi ”mange", (2 sacred mounts: flared), (1*) h: rites and (5) lage chief. The vi the Kapsiki . 3f diSputes, def-333' for ti: EPOResm f0j Often in dip the chief, Becas thief is of: 25" inef is Properly in 3'f-Oéode, he ‘1 the tire-3e 8.] be a _ I °" his . ‘M‘ea w. '10:; ‘village of his father. If a person does move away, including women ‘who marry outside, he retains strong natal village ties, going back in times of family crisis, maintaining at least vestiges of his village "language" and eventually being returned there for burial. Each village is characterized as (1) having a particular "language", (2) having its own set of clans or lineages, (3) having a sacred mountain (here sacrifices are offered and often judgements ren- dered), (h) having its own form.11] ' { {dz eve } { [ Ther. 1133.1 VOWel I no“ Phrases 30" (Cf. 2.2 L. related and the Sew 93'4" , L #8 lb“ in the environment /CY;/2 is represented by /i/ “i, a“ and “u“ /elseswhere/ remain unchanged. For example: {c8} "house" plus {nya} "this" becomes /cinya/l "this house" {dzeve} "hand" plus {nyi} "these" becomes /dzevinya/l "these hands" {t1u} "name" plus {nya} "thid'becomes /tlinya/l "this name" [xu] "flour" plus [nya] "this" remains /xumya/ "this flour" {Ealé} "god" plus {nya} "this" remains /§alénya/ "this god" 2.1.5.3 Suffixes with initial /a/. There are a number of constructions in Kapsiki in which a word final vowel may be juxtaposed to an /a/. These constructions include; noun phrases (NP) consisting of a nominal plus Egg "the thing referred to" (cf. 2.2.5.h), NP's consisting of a nominal head plus another nom- inal related to it by the masculine genitival particle (cf.,2.2.1.1b) and the several verb stems consisting of a verb root plus an extension suffix containing an initial /a/ (cf.,5.l.2.l, 3.1.2.2). The morphophonemic changes resulting from these constructions may be summarized as follows: (1) aIIVf_Il/ya/ (2) V°a||_ _I|/a/ (5) ua“CY_“/uwa/ (h) uaHelsewhereH/u/ (1) Rule one indicates that “a“ coming immediately after a front vowel is articulated /ya/. For Example: {mes} "friend" plus [Egg] becomes /mc€ya§é/l "that friend" {s1} "grand parents" plus {ass} becomes /§iya§€/ "those grand- Parents" [cs] "house" plus {gké} "his" takes the genitive particle {a}, \Vle a gkq| and becomes /csyagké/ "his house" (2) Rule two indicates that the sequence, central vowel plus le.,footnote l on page #7 (section 2.1.5.1). 2‘2: stands for any velar consonant. The notation is read as, /u/'in this, phonemic environment, that is, preceded by any velar con- sohunt becomes /i/. /a/ is repres- i‘inal vowel 0 [2a] "man" [$11) "head Eu a 51:5) and [“1] "I101, ()4) i the (u/ is now For ample: {tin} "ha; 1117.1 a "é” be: bnmfi] Ht] 3.1. 4 Tone 8.: Tone A 1A ~\i ~m~ ~. D . Not a, msic’n Outli. .1 ‘ the notat M9 /a/ is represented phonemically by a single vowel /a/. That is, the final vowel of the preceding word is assimilated to /a/. For example: {2a) "man" plus (gké) "her" takes the genitive particle {a}, “za a gké” and becomes /zagkt/ "her husband". {kagkal "chicken" plus {ate} becomes /kagka§é/ "that chicken" {kelepé} "fish" plus {ass} becomes /kelepasé/ "that fish" {sé} "come" plus {-awa] becomes /sawa/ "come down" (5) Rule three indicates that the sequence /u/ plus /a/, when the /u/ is preceded by a velar consonant, is articulated as /uwa/. For example: [gu] "head" plus {gké} "his" takes the genitive particle {a} ng a gké“ and becomes /guwagk€/ "his head" {xu} "flour" plus {asél becomes /xuwa§€/ "that flour" (h) Rule four indicates that the sequence /u/ plus /a/, when the /u/is not preceded by a velar consonant, it articulated as /u/. For example: {tlul "name" plus {d6} "my" takes the genitive particle {a} Htlu a da” becomes /tluda/ fwusfi] "thing" plus {ass} becomes /wusus€/ "that thing" 2.1.h Tone and Stress. Tone in Kapsiki presents some very intriguing analytical pro- blems. Not all have been adequately solved in this thesis.1 This dis- cussion outlines some of these major problems and describes the basis for the notational conventions used herein. Tone does not play as significant a role in Kapsiki as in manyAfrican languages. It is lexical and not grammatical and there are, moreover, only a handful of words which are distinguished by .Efbch differences alone. There are at least two possible hypotheses WIIii—chcan account for the relative insignificance of tone in Kapsiki. The Chadic languages are a branch of the Afro-Asiatic family. (Cf. Greenberg, 1965, pp. h5-h8.) Most of the branches of this family are not composed of tone languages. Chadic presumably acquired its ‘5“ l 1I am in the process of preparing a detailed analysis of Kapsiki tone which, it is hoped, will appear in print shortly. ‘ “ Q 1 coal one load carri 01 turret \I. H. a. 20101? .1 m J. .zve . it}; SO tonal qualities in the relatively recent past as compared with other African language families. This could account for the low functional load carried by tone in Kapsiki. On the other hand, the tonal behavior of the language, could be interpreted as providing evidence that in fact, Kapsiki is a tone language in the process of changing to a stress-intonation type lan- guage. This type of change is not unprecedented. No attempt is made to answer these questions in this thesis. The tone and stress pheno- mena are treated as follows. 2.1.h.l Basic Tone There are only two contrasting pitch levels or tonemes, rela- tive high (H) and low (L). High is marked with an acute accent over the vowel of the syllable carrying the tone and low is left unmarked (except in a few cases where syllabic consonants are marked to dis- tinguish them from consonant clusters or digraphs.) The pitch pattern of roots, articulated in isolation is termed basic tone. A few of the limited number of words distinguished by tone alone are listed below. xé "they" x5 "night za "surpass" za "man, husband" ma "hunger" ma "mother" kaxé "grinding" kaxe "shooting" Vidi "full" Vidi "night" tléné "work" tlené "tooth" xwfilfi "cactus" xwulu "sweat" méxa "old" mexa "burn" A check of two lists of isolated words totaling 285 items, of Vflnich 185 were nouns and 85 verb roots, indicates that about 65 per Ceant are disyllabic, 25 per cent are monosyllabic, 12 per cent are 'Urisyllabic and only four words contained four or more syllables. (IkIaetual usages, of course, the percentage oflonger words increases Sixme verbs are rarely used without affixes.) The tone patterns for the above sample were as follows: Nearly (”fie half of the disyllabic words exhibit the pattern LH, only one out 0? eight has HL, one quarter has HH and the other one-eighth LL. Of 51 the other polysyllabic words or stems, four-fifths.had high tone on the last syllable. The monosyllables were about evenly split between low and high with nouns tending to be low and the verbs high. There- fore, two out of four of these words, elicited in isolation, exhibit final high tone, a fact of some significance in understanding tonal behavior. 2.1.h.2 Tone Changes. In constructions the basic tone patterns are often altered. The changes are of two types--that induced through affixation and that resulting from phrase or clause formation. (a) Tone and affixation. Some affixes exhibit the same tone pattern with all roots. Other change according to the basic tone of the root. On the other hand, the process of affixation may result in changes in the tone patterns of the root. The patterns resulting from affixation are termed secondary tone patterns. The tone changes occasioned in affixation are illustrated in the examples included throughout the grammatical sketch. Apart from the changes discussed in (b) below, the verbal prefixes (ké:, ka;, tg;_ and.kfié:) retain their basic tone and do not affect the tone of the root. Verb suffixes are either (1) always high or (2) contrastive, that is, low with high roots and high with low roots. The high suf- fixes are marked with an acute accent when cited in isolation and the changing are unmarked. (Cf., 5.1.2.1, 5.1.2.2). Monosyllabic roots and polysyllabic roots with the basic tone patterns HH or LL retain their basic tone when suffixed.l Polysyllabic roots with basic tone patterns HL become HH with the addition of a suffix. And those with basic patterns LH become LL. The following examples illustrate these changes: mené "do" plus -mté becomes menemté kélé "take" plus -te becomes kéléte mexa "burn" plus -te becomes mexaté 1Only the tone of the last syllable is changed by suffixation. Therefore, polysyllabic words of over two syllables need not be repre- sented here. \F‘ P.) (b) Changes in grammatical constructions. Basic and secondary tone patterns are heard in short utter- ances. However, in longer constructions (of several clauses) basic and secondary tone distinctions tend to disappear. They are replac- ed by a phrase level phenomena. This is realized as a raising of pitch and an increase in intensity on the ultima or penult of the in- tonation phrase and can be termed primary stress. If the phrase ends in (l) a syllable with basic high tone, (2) a syllable with changing tone (some prepositions, for example) or (5) a phrase terminator expletive (cf., 2.2.5.5), this will carry the pri- mary stress. If the final syllable has a non-changing low basic tone (Egg "he, she, it", for example) the primary stress will be realized on the syllable preceding it, that is the phrase penult. Although tone distinctions, as indicated, tend to be obliter- ated in constructions of any length, if the intonational phrase is long (over 5 or 6 syllables) a non-final syllable having basic high tone may be articulated at a higher pitch level than the rest of the utterance. This can be termed secondary stress. In summary, relative pitch must be considered at three levels. On roots as a phonemic phenomenon, on stems, that is, on roots with their affixes, and in combination with intensity, on grammatical con- structions. The following examples illustrate some of these features. Key: Dashes indicate basic tone patterns or secondary tone patterns. 2 Superscripts ( ’ ) mark primary and secondary stress. Horizontal, upward-curving lines indicate intonational contours or pitch levels in constructions. Glosses under the Kapsiki transcriptions are literal. Glosses in quotes are idiomatic. male [- -] "Woman" woman male ta gkel [- - - -] "His wife." woman of him lNouns serving as heads of genitival noun phrases behave tonally as suffixed roots. That is, the LH basic tone pattern be- comes LL. tives o 'a male kese [ - ] "The woman came." did woman come 'a male ta gke kese [- - - - -] "His wife came." did woman of him come male kadziyi g8 a nci [- - - - - - - -] "The woman is woman-ing go house of them "The woman is going to their house." naxe dugwu kenefe pa nta xu kwa [- - - ----- - - -] finish pot boil then give flour in "When the pot boiled she put flour in (it)." The following illustrate the use of phrase terminator exple- They are underlined. pa ‘mbeli xkaks HE. [- - - - -] "Then someone called then people called (her)." nda se 'ya pi kanege da nde [- - - - - - - - -] when come I -ing wait I he "When I came he was waiting for me." 51+ 2.2 WORD CLASSES 2.2.0 Introduction. This and the following section on construction classes are concerned with the identification of and distribution of morphemes. For the most part it deals with grammar. However, the relationship between these morphological constructions and the underlying semo- logical structures of which they are realizations will be discussed in a later section. Words are defined, albeit imprecisely, sufficiently for the purposes of this discussion as any utterance which can occur in iso- lation, i.e., in the environment /#_#/. Thus words may be roots-— simple and reduplicated--or roots with affixes. Since there are very few derivational affixes and even fewer inflectional affixes, most words are simple roots consisting of one morpheme. The following classification is an attempt to indicate basic or native categories reflected by the kinds of elements which can be chosen to fill vari- ous slots in grammatical constructions. Thus, the criteria for this classification are the use of words in longer constructions, i.e., phrases, clauses, sentences and discourses. in very general terms five major classes of words can be de- fined, nominals, verbs, qualifiers, relators and expletives. In terms of function, nominals normally serve as the realizations of agents, recipients or beneficiaries of action events in the semology, 01‘ in descriptive clauses which have no verb, as foci of the descrip- tion, that is, as realizations of non-action event sememes. Quali- fiers serve to specify or otherwise modify the nature of an action or t0 indicate the qualities of a noun referent. Verbs are used to realize action sememes and in a limited number of cases to depict states of existence. Relators are of many different kinds and Oper- ate at several levels. They serveto connect elements of grammatical con3131‘uctions either to each other or to non-linguistic entities in some manner. Expletives constitute a special phenomenon in that they do not pattern in larger constructions like other words. Expletives must be distinguished from ideophones (although ideophones may be ex- p IEtiVes) in that many of the common and regularly used qualifiers 55 are ideophones. The above is a functional sketch of the words used in Kapsiki. Following is a more detailed sub-classification of each major category based on formal criteria. The discussion of each category includes(l) a description and (2) a sketch of the distribution of the words. 2 .2 cl NominuSo There are two main types of nominals, nouns (n) and pronouns (pr). Although pronouns serve as noun substitutes and thus fill the same grammatical slots they differ from nouns in exhibiting formal case distinctions. 2 .2 01.1 Nouns a. Number Number is usually not indicated by a change in noun morphology.- If number is to be indicated in nouns it may be done: (1) by the use of a numeral, (2) by the presence of a plural demonstrative, (5) with a plural qualifier, or (A) in rare cases, by the use of y§_which is placed directly before the noun. Examples: (Numbering of examples refers to numbers direcly above.) (1) dzlyi ya vare cs maxkéné Go past three houses. (WS268) go you past house three fig, mcéf te gkwa aéé Five houses are there. (POM) house five on place that (2) EEEEE gyi nza berete mbé... Sticks which are strong. stick which exist strong in (SP5) (5) kwélékwelé kelepé kédza 36... Many fish went. (WS285) many-many fish go themselves kégka va 'ya béxwé I have some chickens. (WSl9h) chicken are I little (h) taga s ezfi le 8 wufé There is grass and trees. 2.. 531... 31...... there grass with tree (PSh) A limited number of words have plural and singular counterparts with the plurals, in some cases, appearing to exhibit the vestiges of a plural suffix. Most nouns with singular and plural pairs refer to classes of people or animals. The following list is nearly exhaustive of plurals used in Mogode. za man Esli men malé woman miyi women 56 wuzegé child ([1 )walc children kwe goat kwi (rare) goats tému sheep timati sheep (pl.) kagka chicken kagkeli chickens mciya bull mcimciyali bulls wusfi thing wusi things méva slave maveEi slaves mentibu guest mentibu§i guests féléxwé seed féléxwéhi seeds Eefa billy goat zefélsmfi billy goats game ram gamélsmfi rams gulu cock gulsmfi cocks zsgwfi kid, lamb hsgwulemfi kids, lambs Furthermore, there are several qualifiers which occasionally serve as nominal replacives also, and which have plural forms. Whether used as replacives or qualifiers theywdll.take the plural form if the noun being replaced or modified is plural. For example: kweté a certain kweci some nya, nyaga this, that nyi, nyigi these, those b. Sex-gender Although the sex-gender system into which nouns can be clas- sified appears to be vestigal, it does evidence itself when nouns are used in a genitival construction. In this type of construction the noun head is connected to its dependent nominal by the genitive par- ticle, which takes the form /ta/ after feminine nouns and /a/ after masculine. Every noun is assigned to one of the two gender classes. Examples: malé woman 8 male ta ca My wife. xwa knife xwa ta Ysgé John's knife. cs house cs a gkél His house. gwalé children gwale a ncil Their Children. Since membership in a gender class does not affect either the form or the distribution of nouns, apart from their use in genitival noun phrases, no indication as to the gender of a noun will be given in the following discussions. lThe morphophonemic changes resulting in the formation of noun phrases using the masculine genitival allomorph, /a/, are described in 2.1.5.5. A noun serving as the head of a NPgen behaves tonally and morphophonemically as a suffixed root. 57 A sample of feminine nouns (this seems to be the smaller class) includes: xa_ "millet", Eéié "woman", dggvg "hand", .Eéié "mush", Egg "goat", gggé "word", melemé "village", Elfiflé. "work", £3313 "penis" Some of the masculine nouns include: pifi' "yams", 29‘ "house", 53 "man", {Qgg "friend", E222 "heart", 31;. "meat", nEi "eyes". Mbst verbal nouns, that is, verb stems used in nominal posi- tions (cf., 2.2.1.1f below), take the pa_ particle when used in NPgen constructions. For example, se ta da "my coming" or pa ta gké "Its price". Words that can stand as a head in a genitive noun phrase are assigned to a noun class designated.by N1. c. Noun Sub-classes There are a number of words which can be considered nouns even though they don't belong to the class Nl discussed above. These may be assigned to one of four noun sub-classes: Locative nouns, Time nouns, Interrogatives and Verbal nouns. Locative nouns (N2) Any locative preposition may take the nominalizing suffixes :52’and.;k§, to form an independent, locative noun, specifying farness or nearness respectively.. These are assigned to the noun sub-class N3. For example: 38 .mcéf t§_ gkwa asé There are five houses there. (POh) house five on place that se teke 'ya mbeda ni... I came here yesterday... (WS206) come here I yesterday kasiyi nde tagé kwa xwégkwa She is there coming in the road. come she there in road pelé "up" pelega "up there" pelské "up here" kwa "in" kwaga "in there" kwaké "in here" laka "by" lakaga "by there" lakake "by here" N nouns are only used to fill the place (PL) slot in clauses. 2 They never serve as subjects of clauses or heads of noun phrases and 58 are analyzed as nouns primarily because the fill the same slots as loca- tive noun phrases. Time nouns (N5) A number of words can fill the time slot (T) in clauses. Since these slots can also be filled with noun phrases and since these words can serve as dependent nominals in genitive noun phrases, they are con- sidered to be nouns and are assigned to the sub-class N5. Like N2's they do not ordinarily occur as subject of verbs nor as heads of noun phrases . Examples: tsé, tsétsé, tsétsénya "now", beEi "today", xénde "tomorrow", mbéda "yesterday", xési "in a short time", mehina "a short time ago", xéca "a long time ago." sé xéndg Come tomorrow! mbé. mbe melemé tsétsépya He is in the village now. mbéda nsexe geg'yé teké Yesterday we arrived here. xetwsd'i nsexe gelj'yé taga In the evening we arrived there. gena ta xéca A story of olden times. _I_nterrogatives (Nu) Interrogative words pattern as objects of verbs and are most easily analyzed as nouns with very restricted distribution. They are assigned to the sub-class Nh' They can only occur as objects of in- txerrogative clauses, never as subjects nor as heads of noun phrases and Hnlst be distinguished from the interrogative expletives occurring at tflie ends of interrogative clauses. The most commonly used interroga- tive nouns are: ‘wa "wha " (also, wa nde "who", we wusfi "which"), tsema "how", tema "where", xwégkwa "when", wareberé "why", gkwini "how many, how much". Possibly the forms ending with.;ma should be analyzed as com- Plex words with :_m_a_ an interrogative suffix. Occasionally it is used With other words to give an interrogative sense, as in: kadze ma_ dze ma ye waa Where are you (pl) going? kwa nya ma gwelepé skeyi mpeldi In which pot should I pour the 'ya waa beer? Other examples of Nu, in constructions, include: 16.: yandeke ga nde waa What is he giving you? 59 wareberé dziyi taga waa Why did he go there? gkwini temfi ve waa How many sheep does he have? xwégfla sé nde was When did he come? ic_e_m__a_ yite a 96. was Where is your father? w_a. mg kwa kwaciga waa What is in the pot? Verbal Nouns (115)l Virtually all un-inflected verb stems can serve as nouns, that is, they can stand as head of noun phrases Examples: (cf., 2.2.l.lb, page 57, for two examples) newe ta Ijké kanewe dé. k1 Look at the look he is giv- "— "— ing you! (8851) 2 .2 .1.2 Prounouns a . Personal Prounouns The personal pronouns can be arranged in the following para- digms. The personal pronouns function as subjects, objects, indirect objects of verbs and as dependent nominals in NPgen. In the first function they take the form of the nominative case. With the exception of the third person, where special forms are used in indirect object and genitive constructions, the oblique form is used for the latter three functions. Nominative Oblique l sing. 'ya dé. 2 sing. na 96. 5 sing. ¢, nde‘2 95, {Jké2 dual mu (gebu)3 bu lThe assignment of verbal nouns to a noun sub-class, N , is made Primarily for the sake of easy reference. In fact, uninflecte verb stems can be considered to belong to two form classes-«N5 and Verb. 2The third singular pronoun is normally realized phonemically 8'8 Zero. However, on occasion the alternative forms are used, parti- cularly to avoid ambiguity. 1"The forms in parentheses are the emphatic forms and are in- cluded in this paradign because with the dual and first person plural ey are sometimes used in non-emphatic contexts. 60 l plural (incl) bumfi (gebumfi) bumfi l plural (excl) n'yé (seo'yé) o'yé 2 plural ys gays 5 plural xé xé ,The genitive and indirect object forms are the same as the ob- lique except as follows: 5 singular, genitive gké "of him" 5 singular, ind. Ob. nda "to, for him" 5 plural, genitive nci "of them" 5 plural, ind. Ob. nda (Object) ke x5 "to, for them" Examples of Personal Pronoun constructions: Subjects 'a :yé képaké maruré I bought rice. (WSl90) kasé x§_xénds They will come tomorrow. (WS202) 'a 292 sé teke He should come here. 'a geg'yfi kédexwfi tli We are eating meat. (WS5l6) Direct Objects 'a na kémpavé gg You hit me. (WS520) 'a kémpavé (Both sub. and obj. morphemes are realized as /¢/.) He hit him. (WS52l) kade 22 mbeli nyi (dual) These people like us. (WS585) Indirect Objects nta gé 'wfi‘we Give me money! (WSSlB) tlene yamené Egg nde He is working for me. (WZlh) pa xé viyi kweté wundu kapa nda ke xé Then they put a certain person to sell (it) for them. (PJh) genaké ca Timatiyé ka geze nda gena ta kanferanse Timothy sent me to tell him about the conferance. The verb geze "tell, talk, say" (see example directly above) and some other verbs of saying, regularly take the nda Object even “when the sense is not clearly that of an Indirect Object. For example: geze nda biyitsa Say it again! (WSl96) b- Emphatic Personal Pronouns With the exceptions of the first and second persons singular, the emphatic form of the personal pronouns is formed by prefixing 5e; tc"the object pronouns. The emphatic of the first person singular is .2fi:z§, that of the second person singular, 52' The emphatic pronoun, 61 whether serving as subject or object, precedes the verb in construc- tions. Following is a paradigm of the emphatic pronoun. sing; dual plural 1 person 'i'ya gebu gebumfi (Inclusive) geg'yé (Exclusive) 2 person ga ge'yé 5 person gegké genci Following are several contrasting examples of the use of em- phatic and normal personal pronouns. .52 kapa xwa YOU are buying a knife. (PZl) xwa yapa‘na You are buying a knife. _'_:_i._'_y_é._ kade kadziyi sire I wand to go to Siir. (TSl) kade L12 katli da teke... I want to leave here... (PZ5) Notice also: sayi 5a 'yé nza 1e Only YOU will I stay with.(PZ5) gggké nde mene tlené HE's the one who works. (SS5h) 5235; yageze nda 'ya IT's this I am saying. (N81) 53 nde meneté nya It's YOU who did this. (wsu12) c:. Possesive Pronouns: dependent and independent. The dependent possesive pronoun is simply the object pronoun 11sed in the genitival construction discussed in 2.2.l.lb above. For example: 52 "head", a da "my head"; xwé_"knife", xwa ta nci "their Iknife". The independent possessive personal pronoun is formed by pre- fixing /ga-/ to the object pronoun, as illustrated below. sing. dual plural 1 person gada gabu gabumu (Inclusive) gag'yé (Exclusive) 2 person gaga gagayé 5 person gagké ganci pa wuferské nda nagké Then she hid his for him... hide for him his (PJ2) kaxwe xé le ganci They are running with theirs. run they with theirs (PJh) dziyi 'ya vaké tenuwe gada I went by here to see mine. go I by—here see mine (PZl) 62 d. Reflexive Pronnoun The reflexive pronoun which occurs in the syntactic position VO_, has the form 06, nji (singular and plural respectively), and is used in several ways. It occurs as the object of certain verbs indicating that the recipient of the action is the same as the actor, giving the clause an intransitive sense. 'a kénti§s pg He sat down. (ws25o) sit himself 'a kétede n§_xwa ysmfi It fell into the water. (WS25h) fall itself in water xwa va ndla nji cs During the rains the houses fell. in rain fall self house (Gr69) The reflexive pronoun frequently occusrs after verbs of motion when the subject is in the third person. It serves to make an in- transitive verb transitive. In other than third person constructions the simple object pronoun is normally used in this position. 'a kégere 9;; He stopped. (While talking.) stop himself (WS256) kéjige 'gg nde She returned home. (PSh) return self she siyi teké siyi Hg nde He came by here. (WSST9) come here come self he naxe kwu kést nji kwa naké The mice ran when they saw us. finish mice run selves when see (SP8) g'yé xé us they The reflexive pronoun is often used with the participial form of the verb when it functions as a noun qualifier. ...kégans pg. ...(it) is pretty. (PZS) kéwuba pg cs nya This house is big. (ws2lu) bigs self house this The reflexive pronoun is used after the numeral kwetég "one" to indicate one thing alone. 'a 56 wundfi kapa ndlé kwetég flé'wé One person alone is not able able person pile wall one self not an enclosure. (SP1) kwetég g§_ kweté te kwembewale a gké Some one is alone in his one self certain on boat of him boat. (PSh) Apart from the reflexive pronoun discussed here, certain verbs may take as object the noun geva "body" giving the construction a re- 65 flexive sense. For example: 'a gena kénéxéve geva... The news spread.... (W8599) word divide body Some verbs take the noun 52 "head" as an indirect object in reflexive constructions when the beneficiary of the action is the same as the actor. For example: 'a 'ya kémeneté nda ke gE.a da I did it to myself. (WSBl9) I do to head of me e. Indefinate Subjects The nouns mbgli "people", Egg "person" and occasionally, nkwa "placefl are used as indefinate subjects meaning "one". For example: kayide ca mbgli Some one makes fun of me. In much the same way the specifier kweté, kweci "a certain, some" and the demonstratives, although normally qualifiers, serve as noun replacives in some constructions. Since they, however, behave as normal nominal subjects in these cases they warrant no special treat- ment as pronouns. 2.2.2 Verbs Since the verb system will be treated in more detail in section three, only a simple sketch of the main grammatical classes and sub- classes will be presented here. Verbs are ordinarily the morphological realizations of action sememes. The minimal form in which a verb can occur is, as a simple root. To the root may be added any one of a number of derivational suffixes which form an extended stem and which serve to change the sense of the verb in some way. The action expressed by the simple root or extended stem may be specified or qualified in a number of different dimensions. This is accomplished by use of one or a combi- nation of the following: (1) inflectional prefixes, (2) markers, (5) auxiliary verbs, (h) qualifiers or (S) reduplication of the root. In their active forms verbs can function as predicates and as heads of verb phrases (VP). There are at least three non-active forms which the verb may take; (1) the stative, (2) the participle, which is formed by prefixing kg; to the root, and (5) the verbal noun which is the root used to fill a grammatical slot normally taking a nominal. Not all of the verbs occur in all of the active and non-active forms i U) IH (1) IL“! V' \ '2rrn ," L It) n 'v» 3Q a: 3‘ c , »Q‘.: ‘u ran ‘ 6h discussed here. For example, mpg "stay" is only stative, 332a "big" only occurs as a participle. However, no practical purpose would be served in classifying verbs on the basis of this type of differential distribution. Following are a few examples of the various forms verb roots may take. (cf., section three for a full description.) Stative verb forms mba 'ya mbe cs kajaga jaga ta as I'll stay in the house and read. stay I in house read read of me (TSl) nti§s 'ya gena dafa I sat beside the food. (TSh) sit I over mush Verbal nouns nexwene ta da my crying (GrSl) (cf., 2.2.l.lf for other examples.) Participles kéline gé yemo nya This water is cold. (Grl5) colds self water this Most of these verbs can also occur in some of the active forms as well, for example: 'a 'ya kéntiés 96 I sat down. I sit self 2.2.2.1 Class I Verbs (V1). Class I verbs include by far the greatest number of verbs in the language. They share the following characteristics. (1) The roots can be extended by a particular set of extension suffixes. (cf., 5. 1.2 :for a complete inventory.) (2) in complex verbal constructions, verbs following a V1 and standing in a purposive relationship to it will take the form /ka- plus root/. (5) Although in practice most Vl's are used transitively, most of them appear to be neutral in this respect. (h) Roots of class I verbs are used to realize generalized action sememes and are normally imperfective while the addition of an extension suffix, in addition to changing the basic meaning of the root also serves to make the stem perfective. Examples of roots and stems: kamené tlené 'ya I work. do work I 65 'a 'ya kémeneté tlené I worked. I do work 'a 'ya kémenemté tlené I did all of the work. I do work Examples of a V1 plus a second verb in a purposive relation: pa Hunter ndeke ca coda kapa xa Then Hunter gave me money to buy give me money buy millet millet. 'a kégaté kanexwené (He) started to dance. start dance Examples of transitive and intransitive uses of V1: 1 . . p§_ yape nde He 18 bathing. wash wash he 'a kén'wusi (He) laughed. Two minor sub-clases of V1 need to be mentioned. The first in- cludes the verbs v§_"put", kglé "take one thing" and £u_"take many things". These behave like other class I verbs except they occur with a series of extension suffixes which are not ordinarily used with the others. (cf., 5.1.2.5 for illustrations.) The other sub-class includes a small number of roots which are used to realize event sememes the recipients of which are singular or, as the case may be, plural.2 For example, Verbs realizing events with Verbs realizing events with singular recipients. plural recipients. kelé . "take" vu "take" petlé "kill" mpfi "kill" mté "die" badlé "die" 2.2.2.2 Class II Verbs (v2) This Class is composed of verbs of motion. The two main roots are fig "come" and.g£e_"go". Several other verbs of motion share some of the unique characteristics of these two and mustlneincluded in the same class. They include, I; "leave", Eli "leavefi Eeké "come home" and jigs "go home". In addition, 5E2 "run" and dzegga "walk" exhibit some of the prOperties mentioned below. ‘ 1This is a transitive verb with a cognate object giving the <30nstruction an intransitive sense. 2The conditioning here is semological, since on the grammati- cEfil stratum the recipient may be realized as subject or object. 66 These verbs are distinguished by: (l) occurring with a set of extension suffixes different from those found with class I verbs. In general these suffixes indicate the topography over which the action realized by the verb proceeds. (2)‘When followed by a second verb standing in a purposive relaisionship to the first the second will have the form /te- plus root/. (5) These verbs, although basically intran- sitive, sometimes occur with a reflexive object. (4) They are often used in pleonastic constructions. (5) The addition of an extension suffix does not make the stem perfective as with V1. (6) The stative form of V2 is often used as a locative with an extension suffix indi- cating "down", "over", "up" and etc. The following examples illus- trate these properties. Extended Vn roots. ntigu kasamé (She) started to come up. (PSh) start come-up 'a kéhavé mbe cs (He) came out of the house. come-out in house (WS279) ntigfi 'ya kasaté... I started to come up. (T89) start I come-up 35 plus a second verb in the purposive. 'a 'ya kadziyi tela nifi da I am going over there to dig my I go-over dig yam me yams. (GrlO) Intransitive use of V2; kadziyi kwe a§é The goat referred to is passing go-by goat that by. (PSh) kelepé kédzaa 96... Fish went in . . . (WS285) fish go-down selves Pleonastic uses of V0; dziyi va xulu dziyi xwégkwa The road goes by the river. go-by by river go-by road (WSH97) fig. nya fig nde It came about. come this come it Extended stems in the imperfect. kadzi i kwe asé The goat is passing by. (PSh) _____11. The Stative form as a locative relator. wa yamené na dziyi va xulu waa What are you doing over near the what do you go-by by river ? river? (PZh) It is possible that this latter use of class II verbs is the hea . J «l.— but t L11 67 same as the first verb in the pleonastic constructions. The first verb can be viewed as filling the place (PL) slot in the clause as it can in the above example. 2.2.2.5 Class III Verbs (V3). Classes III, IV and V are composed of a very restricted number of verb roots with behavior so unique as to warrant considering them 'pseudo-verbs'. They function as verbs in that they are the realiza- tions of action sememes and in some cases pattern like verbs in clauses but they are never inflected and each exhibits other unique features. Class III includes only the root kg_"say". he has the follow- ing allomorph: /pa/ used with first person singular pronoun subjects, and with second person singular subjects which are themselves realized as /¢/ - The main function of the verb is to serve as a quotation marker. It comes at the beginning, at the end or both at the beginning and end of a quotation. Occasionally it is used in other clauses simply as the verb "say". As a quotation marker it may be preceded by several particles, ma "introducer", at the beginning of a quote and, waga, 3313 or ;a_at the end. For example: ma ké kaggwedu sa Pugu ni (quotation) 'a ke xé (NTl) said Mogodians to Pungu, ". . . ." said they kayi wa'a ké Deli Dlexwé "Kai!" said Deli Dlexwe. (NTl) As illustrated here, the recipient of the action of kg is marked by the preposition Ea. .§2 is the regular preposition indicating "ac- tion toward" used after verbs of Class II. When the agent of the verb or the recipient of the action is represented by a personal pronoun, certain changes take place in the form of the verb, i.e., the allomorph discussed above is used. The following examples illustrate these changes: pa 'ya sa ga I said to you. say I to you(sg.) pa 'ya E6 I said to him. say I to pa 'ya ta xé, or ...sa genci I said to them. say I to them to them(emph.) 68 pa You(sg.) said. pa so You (sg.) said to him. say to ke sa ga He said to you. say to you(emph.) ke xé Ea 'i'ya They said to me. say they to I(emph.) ke mu Ea gala We say to god. say we to god wawa ké Ea gebumu He said to us all. thus say to us "...‘ya kadema zele gena ass" pa "...I don't believe that word," I not believe word that say you say. (SB2) ke mbeli sa geze One says. say person to saying pa 'ya sa geze I say. say I to saying 2.2.2.l Class IV Verb (V2). The Class IV verb, ya, has two allophones, /va~ve/ and like ‘kg is uninflected. It is used to show possession, as a locative verb and on occasion indicates action toward. (Like V:5 and V5, which also show phonological affinities to prepositions, this verb may be etymologically related to the preposition Ea_"near", "by".) The z§_form of the verb is found in final position, that is, before the ¢ realization of the third singular pronoun. (Notice in the examples directly above the change from §a_to E2 in the preposi- tion when used in this same environment.) Examples : gkwagé kagkeli va_'ya I have six chickens. (WS558) six chickens I ...gké keleté ca lekwesa ta ca ...he took my shirt to John for he take me shirt of me me. (TSl) XE. Zagé to John pa 'ya keleté wusu .Xé Then I took the thing he has. then I take thing has (TSll) Occasionally the particle ya_is prefixed to the verb. This same particle is sometimes used before prepositions to indicate loca- tion. The use with 12 appears to be entirely optional. 69 Examples: degwe ménéfé yavé He has a new pet. (WShS5) pot new is to him zsgwe yavé ki gs She has a kid in the house. kid is in house (PEA) The verb xa_should be distinguished from the relator 12 "be- cause" (cf., 2.2.h.5). Again the two words may be etymologically re- lated but the relator is found in very different grammatical environ- ments, for example: gala X8 kelsteti kambsxwé XE The dogs barked all night be- all night dogs barked because cause of the hyenas. (SP2) degwava gwe. hyeana Eswe 'ya va ma I am hungry. (WS #29) hurt I because hunger 2.2.2.5 Class v Verb(V5). The pseudo-verb §a_is used as a verb replacive. It may be re- lated to the preposition Ea which is used with V2 to indicate action toward something. As a verb or verb replacive §a_ is used in construc- tions with an ideophone which enjoys a semantic relationship to the verb being replaced. For example, léng and pgéké are ideophones both used to qualify the verb kglé "take". (Cf., 2.2.5.1 for a description of ideophones.) They are commonly used in clauses with the verb, as, for example: nzéké keleté telsya Suddenly (he) took the cache- pow! take cache-sexe sexes. (NMl) Very often, however, the verb is ommitted in such constructions and the ideophone is used with the pseudo-verb Ea, For example: nzaké Ea ysmu (He) suddenly took water. (NMl) pow! water lage E2 geta a mega a g'yé (He) grabbed the stick of our chief. suddenly stick of chief our (NTl) daxé 'yé. E2 xsdi Immediately I picked up dirt. (NNl) P0P: I dirt gd'a [Ea ts wuga (He) prepared beer then. (NTl) beer then mbiké §§_ livu gké (She) tied on her apron. (FMl) apron of her 70 2.2.5 Qualifiers. The qualifiers, in terms of their occurence in constructions, fall into five sub-classes. They are (l) ideophones, (2) numerals, (5) the specifier, (h) demonstratives and (5) verb markers. Qualifiers occur in constructions with either nominals or verbs. 'When used to modify nominals they normally occur in the syntactic slot _N if the nominal precedes the verb and in the position N;_ when the nominal follows the verb of the clause. The qualifiers do not constitute a mutually exclusive class in every environment. For example, verbs can on occasion serve to modi- fy other verbs and qualifiers (notably the specifier) sometimes func- tion as noun or pronoun replacives in verbal constructions. Further- more, nominals may be modified by verbs occurring; in an active reflex- ive construction, or the stative form tied to the nominal by the re- lators kwa or Eya. For example, ysmfi nya liné Cold water. (Gr5) water this colds kéliné gé ysmfi Cold water. (Gr5) cooled self water gggské ndel It is pretty. (PJS) goods it 'a kékeleké kwaciga {g§fl§_ (He) brought a pretty basket. (wshhh) brought basket good ...kégags 98 (It) is pretty. (PJS) goods self A number of qualifiers always occur as reduplicated froms. For example; kwélékwelé "many",'much". In addition, both ideophones and numerals can be reduplicated to indicate intensity, continuous ac- tion, or, in the case of numerals, enumeration. For example: kadzegwa wuzege nyaga ntar ntar ntaré That child kept walking walks child that like a toddler. (FM5) ...xa maxkéné bak bak ysdla ...three stalks of corn with two corn three two two ears ears each. (PO ) lSince gaps and several other verbs that function almost ex- clusively as qualifiers occur in the slot ordinarily filled by ideo- phones, it would be possible to assign them to a sub-class of qualifi- ers. Hewever, since these can all occur in verbal constructions and with verb affixes, it is more efficient to consider them as verbs with a peculiar distribution, analagous to that of ideophones. 71 2.2.5.1 Ideophones. Ideophones, which are common in African languages, have been amply discussed in the literature (cf., Stennes 1967. 17-20. Samarin 1965. et. al.). Ideophones in Kapsiki exhibit several characteristics not common to other words. (1) the class as a whole is productive, that is, new and sometimes idiolectical forms are continually being formed. (2) Many ideophones are onomatopoeic. (5) Along with numerals, ideophones reflect the presence of a phono-tactic sub-system which re- sults in tone and consonant-vowel patterns not found elsewhere in the language. For example, pitch of ideophones is often higher than nor- mal high and final consonants are common. (Cf., 2. 1.2.ha above) Ideophones themselves are of two types. The first can be termed adverbs, since they function exclusively as verb modifiers or verb substitutes. The others can serve as nominal, verb or clause modifiers. a. Adverbs Adverbs ordinarily occur in the syntactic environment #_SV or more rarely, #_VS. They may modify V1, V2 or V5. Most adverbs evi- dence a semantic identity with verb stems or classes of verb roots they modify, reflecting an’ordered downward and' relationship between the semological and grammatical strata. It is this quality which per- mits adverbs to serve as verb substitutes and to modify the pseudo-verb E2?) (Cf. discussion and examples in 2.2.2.5 above.) Adverbs are often onomatopoeic and when qualifying verbs de- picting repeated or continuing activity may be reduplicated. Follow- ing are some examples of adverbs, first, with verbs, then serving as verb substitutes and finally with fig. ' pékwé késé nde Suddenly she came. (FM5) appear! come she tséxpetlu tséxpétlfi tséxpétlfi nde kadziyi as She went walking (imitation of a frog walking) she went self like a frog. (FMl) nzaké keleté (He) took (it). (FM5) pow! took végfi kelexwu gu Whang! (She) took off his head. whang took-off head (FM5) ntinfil male aSé kazekwfi left woman that return 2 cape kajigs 96 left go home self pélésé Kweji mbele 96 zoom! Kwaji escape self nzaké kwagamba That woman returned. (FM5) (He) returned home. (PPM) Kwaji escaped. (PPM) Kwagamba took it. (FMl) sudtnly Kwagamba daxé Ea xa She picked up the millet. (In a business-like millet way. (FMl) pa'a sa velugfi kwa PlOp! She put the beetles in (it). plop! beetles in (FM5) gwu gwuggwu Sa geta dewé Bang, bang, bang, the stick(hit) the bang bang bang stick wall b. Ideophones (general) Ideophones of the second type may either precede or follow the word or construction being modified. These qualifiers are used to ex- press a wide range of meanings. The sub-class embraces quantifiers, locatives and qualificatives, including most morphemes referring to colors. Some examples follow: mbeli cuk3 All of the people. (ws181) melemé fét3 All of the villages. (WSl82) kwe ta as krép3 goat of me all All of my goats. 'a kéffite kulu pats (He) picked up all of the take piece all pieces. (WSShS) lntigu presents some problems analytically. It is used com- monly with V and often with what seems to be the purposive form of the verb, as above. This suggests that it is a defective verb. This analysis is supported by the rare occurrence of examples where it is used with finite verb affixes. Hewever, the tone pattern, the slot it fills and its frequent occurrences with non-purposive verb forms seems to justify the inclusion of it in the adverb class. The occasional uses as a finite verb can be explained by analogy, since it does en- joy a close semantic relationship with V2's. 2capé, although rarer, appears to be a synonym with ntinu. 3cfik, fét, pét and krépgé) are all synonyms and, insofar as my data indicate, may all be used interchangeably. takwa takwa kwe yité The goat was near (her). (PPh) near near goat beside kwélékwelé sé mbeli kala timéti The people were coming as many-many come people like sheep numerous as sheep. (SP5) pa za a 916 tsaxeké kwélékwelé Then her husband thanked her man of her thank much-much very much. (PZ5) 'a kékeleké kwe dibé dibé (He) brought a fat goat. take-here goat fat fat (WSth) gémé gémé pels nyi taga These stones are very red. red red stone these on-there (WSh60) 2.2.5.2 Numerals. As already indicated, the numerals in Kapsiki behave like the ideophones in many respects. When used to modify a nominal the cardi- nals may either precede or follow the nominal, depending on the posi- tion of the verb in the clause. For example: bak mbeli or mbeli bak Two people. mcéf kwaciga or kwaciga mcéf Five baskets. The numbers from 1-10 are realized as simple morphemes while the higher cardinals are poly-morphemic. kwetsg one gkwag(é) six bak(é)l two mberefég seven maxkén(é) three degesé eight. wufad(é) four metli nine mcéf five meg(é) ten meg kwetég gumbe eleven ten one meg bak gumbe twelve ten two bake msu twenty two mcéfé msu fifty five I cannot positively identify the meanings represented by the 1The final phonemes which are enclosed in parentheses, repre- sent allomorphs which appear to occur in free variation in the simple forms and in some to the combined forms as obligatory. 7% particles ggmbg and Egg since they don't occur elsewhere in my data. There is in Kapsiki a term for one hundred, gumsfikfi but it is rarely used. The Fulfulde term temgrg being more common. The word for one thousand is the word for 'Sack'. This parallels the Fulfulde usage although either the Fulfulde word, Egrx§_or the Kapsiki word tlibé may be heard. Ordinals are composed of the cardinal connected to a preceding nominal by the kwa relator, followed by the particle Eggg "times".l Thus: za kwa bak zege The second man. (SP1) man rel. 2 times "first", however, is expressed by using the verb gaté "start, begin" connected to the preceding nominal by the relator kwa. za kwa gaté The first man (SP1) man start 2.2.5.5 Specifier. The specifier, kweté (Sing.), kweci (p1.)'h.certain, some", may modify any noun, function as a noun—SubStitute and incidently serve as a number marker. For example: wa nyi kweci nyi mbeli teké kaxé kelepe ta nci gi Here now certain these people here fishing fish of them(excl.) kwetég gs kweté te kwembewele a gké te nduru. maxkéné one himself other on boat of him on lake three kweci pelské. kweté biyitsa te kema ta xulu (PS5) others up-on another still on face of river Here are some people fishing. One is by himself in his boat on the lake. Three others are up here.~ Still another is in front of the river. ntigfi kweté wuia makwa kadzemté A certain girl left to go to left certain girl go-to-bush to the bush. (P01) ’3 kweci mbeli kenaké Some people saw him. (WSl92) some peOple saw 2 .2 .5.ll Demonstratives. The demonstratives, like the specifier, have singular and non- Sfiiiigular forms. There are three sets of demonstratives, gyg, Eli \ lPossibly this should be analyzed as a verbal clause. Egg is usu-a.1ly used to relate verbs to a nominal. (cf., 2.2 .b. .2) Thus Esgé c30mm be considered a verb with limited distribution. 75 "this", "that", pyaga, nyigi "these", "those" and ahél "the one refer- red to". In addition to their function as nominal modifiers, the near demonstratives can also serve as relators (cf., 2.2.4.2 below.) The near and far demonstratives may either follow or precede the nouns they modify, but they ordinarily follow, while a§€ always follows. Examples: kéwuba gs cs pyagé That house is big. (WS2l9) bigs self house that nya xwa ta as This is my knife. (WS2l2) this knife of me wa wusu pya 'waa What is this? (WS2l5) what thing this ? baké gyigi mbe dzeve va wundu aSé (PS5) two those in hand of person referred to Those two are in the hand of that person. wa tlené nya mené miyi gyi ki gs a nci waa (POl) what work this do women these in house of them ? What work do these women do in their houses? wa meleme ass waa What village is that? (F07) what village referred to? wundu §§§_ ndé ggwedu That person was ggwedu. (FC7) person that one ggwedu 'a so kave nda xa gags ke male a§é wé (FC8) able give it millet good to woman referred to not He didn't give good millet to that woman. mené nde kala EIEEE She did like that. (FC8) do she like that wuba ma cs gya ké nya This house is bigger than this. big -er house this to this (SS5) As illustrated in the last two examples above, the demonstra- tiVes can also occur as replacives for nouns or longer constructions. hfluerl used in this manner the antecedent must occur in the immediate en- Vir Onment . The phonological sequence /nya/ also occurs frequently as a par- ticfilxe suffixed to words of other classes. Whether this is the same \ 1 1Although written disjunctively here, fléé can possibly be ana- yzed as a noun suffix since (1) it always follows the noun directly, and (2) it causes the phonological changes described for suffixes in “‘1-3-5 above. 76 morpheme as nya "this" is unclear from my data. For example, tsetse "now", a time word, sometimes occurs as tsetsenya and vanya, vadanya "because", a relator, sometimes occurs as va. 2.2.5.5 Verb Markers. Verb markers, as part of verbal constructions, are dealt with in some detail in section 5.1.5. This section will simply provide an inventory, a statement concerning distribution and several examples. Verb markers are free morphemes which function to in some way specify the action of a verb. They occur only with the active form of verbs and ordinarily occupy the syntactic slot #(time)_SVO(sat)#. The exceptions are the negative marker, wé_and the interrogative markers, Eéé andkyi, which come at the very end of the sentences they serve-to negate or make interrogative. The verb markers are: fig 'active marker'(cf., 5.1.5.1), pa 'con- secutive sequential marker' (cf., 5.1.5.6), nda, 'incompleted sequential marker' (cf., 3'1‘5°9):.§222 'purposive marker' (cf., 5.1.5.10), wé 'negative marker' (cf. 2.5.5.5b), y; and.§é§ 'interrogative indicators' (cf., 2.5.5.5c). Following is an example of each type of marker: ‘Lg kweci mbeli kenaké Some peOple saw him. (WSl92) some people saw p_a_ ndeke da wusu pa dzegwa 13!: 1e xwé (SS29) then gave me thing then walk himself with running Then he gave me something and then ran away. ma nda késé nde nda tezemé 'ya When he comes, I'll eat. when comes he then will-eat I (W851?) cikéké keEa fa kwada... (She) paused to hear the noise... pause for hear noise (PSl) kancs nda geva nde ké mbeli wé She didn't Show herself to the Show body she to people not people. (FM6) kaisne mp1 nde xi. Is he breathing? (W3225 remains breath he ? 2.2.h Relators. Relators are function words which serve to relate grammatical units to one another. They fall into several different distribution classes which are termed: (l) introducer, (2) relatives, (5) preposi- tions, (h) genitive particles, and (5) conjunctions. 77 2.2.h.1 Introducer. The introducer, ma, occurs in the initial slot of clauses and serves simply to introduce the clause. It is used in a number of dif— ferent grammatical contexts. Following are some examples. It is frequently found preceding the verb kg "say" introducing quotations. ma ke xé SC ni "(quote)" They said to him, "...." (FM5) say they to ma ke mbe gu a gké "(quote)" (He) said to himself, "...." say in head of him (FMA) It is sometimes used to introduce stories or fables. ma kweté wundu gé xwa kaggwedfi gé n1. A certain person among certain person among Mogodians the people of Mogode. In introducing the first clause of a compound sentence it may be used with an uninflected verb stem, when there is no necessary tem- poral connection between the actions of the two clauses, or with one of the verb indicators discussed in 5.1.5, below. .22. zeze 'ya ni kameneté 'ya I think I can do it. (WS262) think I do I ma kwamenemté tlené nde ntigu nde After finishing the work he after-do-an.work he left he left. (Gr70) ma_nda tesé va naxe cs kéndla 96 When the rain came the house when come rain did house fall self fell. (Gr70) In a very few cases mg is used with the particle ya to intro- duce an action which didn't eventuate. ma ya kadema va késé nda katendla cs wé (Gr68) if not rain come then fall house not If the rain hadn't come the house would not have fallen. 2.2.4.2 Relatives There are two relatives, pyg and 533. In some grammatical eon- texts they are only distinguished stylistically but in others they are used to express different kinds of relationships. kwa is ordinarily used to relate words or constructions to nouns 'which they qualify. For example: male aSé kwa juni ta za a gké That woman is her husband's woman that loved of man of her beloved. (PSh) lekwesa kwa gkiréyi yave (He) has black cloth. (wshvv) cloth black has I .— 1 C II .If r l . 1 Cu . . 3 r1 0 1 r. r c c e e t e .l t S . b“ on a. .-l 0 JV . -.— 0 .sb m mi 6 wk. AN. 1, g 1! In?“ I or." +1... "1.! Wm :1 c. .n +. we e .vau ..AIL. .m m a: flu Wu. . 1- pew-xv . ier of “y. :31. b O .ni. S V we. a. th ...w SE Y..- r ‘ - ‘eallZeé “In 78 In the terms for "left" and "right" the kwa relative plus a qualifying verb is used as a single lexeme ordinarily filling a syntac- tic slot reserved for nominals. The noun to which kwa might have ori- ginally related the verb (probably dggyg "hand") normally is not used in this construction. For example: ...pa na zsreté geva te kwa zemé ...then you turn to your right. then you turn body on which eat (W8268) wufé dzemé vagé kwa gwelal Trees are to the left of her. tree go beside which (P02) EZE has a much wider distribution than kwa. Like 5E2 it may serve to relate a qualifying word or phrase to a nominal. In the two examples of this construction below, the first illustrates its use as the pleonastic subject of a clause, in which case it parallels the be— havior of the emphatic personal pronouns discussed in 2.2.1.2b above. wundu gya kexwé kaxkaxé ndé g'wuvu.wusu (SP ) person who run fast one wins thing He who runs fast wins. nde nya ketliyaxé 96 He is thin. (Grlh) he who thinned himself nya is also used to relate two clauses. This construction nor- mally takes the following form: S2 nya V1 (0) (I0) nya Sl V2 (sat) Clause2 is imbedded in Clause . Clause also functions as the 1 1 object of V1. If S1 and 82 are realizations of the same sememe, S1 is realized morphemically as ¢. For example: wundu gya geze gya 'ya gsni naxe késé (SP1) person who speak I then! finish come The person about whom I spoke has come. wundu gya naké gya na gegké gké The person who you saw, that's person who saw you HE him him. (SP2) cs gya xé Eya mu mbe naxe kéndé he (SPll) house which slept we in finish burn itself The house which we slept in has burned. In the above illustrated types of construction, the second nya is frequently omitted with no apparent change in meaning. For example: lWhile right can literally be translated "(hand) which eats", the meaning of gwela--if it had one--seems to be lost and it is used with £33 to form a lexeme meaning simply "left". Thus, "left hand" is dzeve kwa gwela J Y the 'CE 9. C OW 20201+°3 the a: ":1 1 rd A; wk. in. o a v . . w a,“ 79 tla gyi geze 'ya nagsé teké xé The cows about which I spoke cow who speak I before here they are here. (SP9) 2.2.h.5 Prepositions. Prepositions serve to relate nominals or noun phrases to claus- es in a type of satelite (sat) construction. They are used to express several different kinds of relationships. The most commonly used pre- positions and the relationships they realize are listed below. (a) Locatives; mpg "in" (as in a hand), kwa "in" (as in a bas- ket), kg "in" (used exclusively for in a house), 13,1223; "beside", £2 "on" (very general), yipé "near", "beside" (as in the vicinity), p212 "up on", laka "over by", xwa "among", kwélé "between", 5222 "over" and £222 "on". (b) Instrumental or accompaniment; 12 "with". (c) Goal; kg "toward". The above glosses must be considered very rough approximations since the prepositions in Kapsiki cover very different segments of the semantic field from those in English. This is true from the standpoint of differences in meanings (see for example, the several prepositions glossed "in" above) and from the standpoint of the function of the class as a whole. In Kapsiki, for example, prepositions do not modify in any way the action expressed by the verb but simply specify the relation- ship of the satelite to the action. Thus one says: 'a kélakaa wuzege pels wufé He took the child down from the take-down child up tree tree. (ws278) 'a kéSave mbe, cs (He) came out of the house. come-out in house (WS279) It is apparent from the above examples that much of the func- tion of English prepositions is assumed in Kapsiki by the verb stems. This will be discussed in some detail in section three below. The several examples that follow illustrate the range of mean- ing, normal distribution and special functions of prepositions. a. Locatives ’a kémenamé hSi mbe zewé He made a knot in the rope. did-in knot in rope (WS545) ...pa 'ya belaxs ca mbe xa ...then I laid down in the corn. then I lie me in corn (TSl) wa wusu nya kwa kwaciga waa ‘What is this in the basket? what thing this in basket ? (wssog) 80 dziyi laka gs da dziyi gé nde He passed my house. (WSSBO) go by house me go himself he kwélé wuve a da melemé wuve a gké (WSSOl) between farm of me town farm of him His farm is between the village and my farm. nde ... katepa tg_§skwu tenza Perhaps he will sell them in he will-sell on marked perhaps the market. (P83) pa dzexwfi gé xwa xa Then it went into the millet. then go itself among millet xwa is also used to express a temporal relationship, meaning "during". For example: xwa va ndla 96 cc During the rain the house fell. during rain fall self house (Gr69) laka is occasionally used to express an instrumental relation- ship in a passive construction (cf., 3.2.h.2). laka gegké ndeke da geva xwa The knife was given to me by ‘ by HIM gave me body knife him. (GrhO) b. Instrumental The preposition lg is used to express a number of types of re- lationships. The most common being instrumental and accompaniment. For example: kamené tlené 'ya lg wudfi I work with a hoe. (Spl5) do work I with hoe 'a kégeze lg. berté (He) spoke loudly. (WS255) spoke with strength I .... kapaké ye kelepé le gwené . . . (she)'will buy fish and buy (pl.) fish'fiith salt salt (PK?) kateta dafa l2} She will cook mush with (it). will-cook mush with (PJR) mcé xé lg_ male asé They are friends. (P29) friend they with woman that ... pa dzegwa 96 lg_ xwé Then (he) ran away. (WSDSO) then walk self with run taga ye guzfi [lg ye wufé There are grass and trees there. on-there grass with tree 1When the preposition relates a nominal which has already been mentioned so that the third singular pronoun would normally be used, this is realized phonemically as /¢/ so that the clause ends with a Preposition. ‘I bl." a «r: Tier-4 , O 81 c. Goal Like lg, the preposition kg covers a wide range of relation- ships. Some of its functions have already been discussed, as its use as an indirec object marker (cf., 2.2.l.2a). kg is used to relate a second entity to a first in a compari- son. jamajama ma cs kaMargi 3.3 kaLemté (SP1) big pass house Kapsiki to Mafa The houses of the Kapsiki are bigger than those of the Mafa. biya ma segwu nya k§_ nya This stick is longer than this. long pass stick this to this (337) kg is used to express the relationship of ownership. wundu. kg kwe taga wé The goat's owner isn't there. person to goat there not (P02) kg is used occasionally to relate a qualifying word (descriptive verb or qualifier) to a nominal. Unlike the relatives, kwa and nya, in this type of construction the qualifier precedes the nominal. line 1_<_é_ yemfi - The water is cold. (wsuss) cold to water yEmfi nya 'yana ‘ké The water is hot. (Grlh) water which hot to kg also functions to relate a nominal to a verb as the goal of the action of the verb. megela nde berté ké’ He is well, he has strength. good he strength to (NMl) nzaké ‘53 Ea kweté wusu He took a certain thing for him- took to certain thing self.) (FMB) 2.2.4.h Genitive Particles. The genitive particles, which have been described in the sec- tion on nominals (cf., 2.2.1.1 above) function to relate nominals to each other in genitival constructions. In these noun phrases, the par- ticle 212a is used to connect a genitival modifier to a preceding nominal. For example: nde kataké keZa xé le za .9 gké le wale a gké (P22) she cooks for they with man of her with children of her She cooks for her husband and her children. 'a kelemté dzeve ta_da (He) pinched my arm (wsssg) pinch hand of me 82 2.2.h.5 Conjunctions. The conjunctions serve to relate a number of different kinds of grammatical units. For the most part they operate at the phrase or clause level. The preposition lg_is the primary form conjoining nomi- nals at the word level. In the following discussion the most import- ant conjunctions are listed, their normal distribution sketched and they are illustrated. a. ama "but" . Although borrowed from Fulfulde, EEE is very commonly used in Kapsiki speech. It ordinarily relates two clauses in a contrastive re- lationship. For example: cikéké keia fa kwada za a gké késé amé.kénaka kasé wé (PZ2) paused for hear noise man of her come but ripe come not (She) paused to listen for the coming of her husband but it was not yet time for him to come. kafiya que ass ama 'a sc kays nda gena wé * (PSh) forbid girl that but able respect it word not That girl tries to stop it, but it won't mind. pa ntléte balé ama kétebfiké wé He jumped but couldn't reach it. jump jump but reach not b . ba'a "and" ba'é is a rarely used coordinating conjunctions. It is op— tionally used in connecting verb clauses denoting simple succession of actions, which are usually marked by the verb marker pa and in a few other cases. gegké sé keEa zeme wusu ba'a keEa mené tlené (Gr75) HE come to eat thing and to do work He came to eat and to work. pa ndeke da cedé ba'a pa geg'yé gezaké gena (GrYS) gave me money and then we talked word He gave me some money and then we talked. c. seyi, dla "only" These two conjunctions are used interchangeably or in combina- tion. EEK; is borrowed from the Fulfulde while dla appears to be the vernacular Kapsiki term. ma kwampavé 'yé seyi nexwené ya nexwené nde (GrY?) after hit I only cry cry he When I hit him he only cried. wusfi thing He dc d. kala J structio: wufé tree mene doe. e0 van 1‘ trated be malé Woma This male worse; That féré ha' Her 1 83 wusu s6 kamené nde wé 58y} dla nza mbe c8 kaxe (er7?) thing able do he not only stay in house sleep He doesn't do anything except stay in his house and sleep. d. kala "like"L "as" kala is used to relate two nominals or other grammatical con- structions as being comparable or equivalent. wufé dziyi taga ... kala maggweré The trees there are like tree go on-there like mangoes mangoes. (PSh) mene nda nde ke za a gké kala nya She does for her husband like does to she to man of her like this this. (PSI) e. vanya, vadanya, tumberekénya, vatse "because" These conjunctions may all be used interchangeable, as illus- trated below. malé nya kagfi miké vanya dzemté za a gké (PSI) Woman this cooks gruel because goes man of her This woman is cooking gruel because her husband is away. male age kwa juni vatsé mene nda nde ... (P81) woman that which love because do it she That woman is the loved one because she does . . . . faréfaré wune a megeyé tumberekéqya xfike nde (P22) happy heart of mother because grind-bring she Her mother is happy because she ground for her. ... adénya nza nde garé gweze as: . . . because there is much because is it much grass that grass there. (POH) 2.2.5 Expletives. Expletives are uninflected words used to get attention, make responses, mark hesitation or to otherwise vocally punctuate speech. As used here, the term expletive includes, interjections, grammatical- ly acceptable exclamations and a number of other words which are not easily classified. In addition to these functions alluded to immediately above, expletives in Kapsiki appear to serve the purpose of providing morph- emes to carry the rising pitch which is typically heard in phrase final position. The presence of an expletive, which adds nothing to the meaning of the phrase, permits the realization of this intonational phenomenon even when the final phrase syllable would normally carry low tone. In some speaking styles the expletives, 213'22; HE; wuga, wuka, 3E, 522;, pg, are heard very commonly as phrase terminators. Examples of the most common expletives are listed below with some a1 task i: listic I! 5.2.5.. tions catch 3.11 (3' H a phys n: he rn‘ — O o 1. (cf ly in Cbres H 8h same attempt made to describe their normal distribution. The latter task is complicated by the fact that expletives represent largely sty- listic and individualistic variations of speaking. 2.2.5.1 Expletives of Attention. In addition to the introducer, ma, which in some rCSPCQtS func- tions like an expletive, there are a number of words which serve to catch a hearer's attention. The most common are: a. nda(a) "here" The optional long vowel probably reflects the fact that this expletive is borrowed from Fulfulde. naaa fa 'ya kéde kamené geva le 'ya g6 (TSl) here hear I wish do body with I then Here then, I'll understand what will happen to me. b. ndé "here" This is often used when offering someone something, either as a physical act or as a matter of speaking, for consideration. ndé dafa Here, eat: (GrSO) here mush ndé mu Come here: (Plural) (Grh8) c. yflwu "O. K. then" This expletive is possibly also borrowed from the Fulfulde (cf., Stennes, 1967. 159). It is often used to start a story, eSpecial- ly in reSponse to a request. wu fa mu ndedekfi Now then, listen well. O.K. then listen well ygwu kala gena nyi ... O. K. then, like these words. (SB?) Frequently the alternative pronunciations, yawaa or XE are heard. This appears to conform more closely to the Fulfulde form of the exclamation. yawaa ma mesina ... O. K., in a little while . . . . xanaké Sunu Good then, thank you, Sunu O.K. thanks Sunu . d. a "Ah" a’is used to introduce statements about which the speaker feels some hesitancy. ll of attent: formal ed written h For exam; gig 2.2.5.2 E.‘ ,4. eating as her of id: the genera a. 'é'é 'é‘é} \ no I 85 'ya kédepé wusu nya ... Ah, I don't know about this I know thing this thing. ng a_kadema g'welu da kageze Ah, my voice isn't working. ah not voice me say Some Speakers, particularly the Christians who have had some formal education in Fulfulde, tend to use Fulfulde words as expletives of attention. Notably among these are xa_or gala (hag in Fulfulde as written by the protestant missions) and kéyi (written kay in Fulfulde). For example: Eéli. malé nya ni ... New: this woman. 2.2.5.2 Expletives of response. There are two commonly used expletives of response, one indi- cating assent and the other disagreement. There are, however, a hump ber of idiolectical variants of these words recorded in my data. Only the general terms will be illustrated here. a. 'a'a "no" 'a'a kénaka 'ya xevé kwa ceracéra No! I haven't even caught no not-ripe I catch which tiny a tiny one. (PZl) b. e, e'é "yes" This expletive, or a variant of it, is used frequently, either as an affirmative response to a question or as an indication of agree- ment in the telling of a story or making of a statement. 'a nda saté na se tsé. e'é Then you have come to his house then came you to nOW’ §E§ now? Yes. (0K2) 2.2.5.5 Phrase Terminators. This rather disparate class of expletives is defined by always terminating a phrase, always having high tone--often higher than nor- mal high--and, as with the other expletives, exhibiting a great deal of idiolectical variation. The most common and generally used are il- lustrated below. a. n1 ~ né 2;, which is sometimes heard as he, is always pronounced with a'very'high tone and indicates that something more is coming after it. It is the most common of the phrase terminators and is frequently used in combination with others. 86 kadzegwa xé te ysmfi n; kakesé ... They are going on the water go-on they on water to-catch to catch . . . (PZl) ma ke §i_ kégavé ... She said, "if I give . . . ." said give (PJ5) ma zeze 'ya n; kameneté 'ya I think I can do it. (WS262) think I do I tfimfi veci n; kazeme wusu 'ya I eat every day. (SSH) every day eat thing I kézané yita gu a da gsni forget beside head me then ma gele ya 'ya nagtse tseni scold you I recently then ma gena nya banda né_... word this thus then b. gs (Sometimes heard as ga, ka) I have forgotten it, then. (NK16) I scolded you a few minutes ago, then. (NY2) This word then . . . . (NY2) This expletive, like pi, occurs by itself or in combination with other morphemes, for example, geni (cf., directly above) wfigé, wfika, and g6 . high rising tone. and may have the force of making the clause a rhetorical question. In the combined forms the final tone is normally a The simple form is ordinarily used clause finally The combinations with.w§ always signal a question, either rhetorical or one demanding a response. ... kadzaté gs a nci go house them sé ndé sfigkwe ga gg here money you then ... gwalé va xé géwfi children have they then vatameda g'yé wuka last-year we then c. wfi . . . going to their house. (Psu) Here is your money, then. (CP2) . . . they have children then? (CP2) Last year we (fished) like that? (NY2) The expletive wfi can also occur in a number of combined forms (cf., the examples directly above). It is used either as gé_simply to end a phrase or to give a clause the force of a question. 'ya laté Geru wuga I up Guria then ndedeku gé kanasaré pérfi LEE good self Whites also then " ... " ké tsewuka say then When I was up to Guria? (CP2) Whites are also good, isn't that so? (CP2) ". . . ." he said, didn't he? (NY2) 87 d. ki The expletive k; is normally used as an intensifier with the imperative but occasionally occurs with other types of constructions. gagé geys kaggwedfi ‘5; You people of Mogode are very good YOU people of Mogode good. (NY2) zemaké 5; ' Eat a little: (GrSl) newe geregé ta gwalényi k; Look at how those children are look stand of children those standing! (Gr51) ndeme kiki Taste, please! (Gr5l) 2.2.5.5 Exclamations. The expletives in this sub-class do not function in any of the specific ways mentioned above. Again, my data reveal a great deal of stylistic variation in usage. a. banda The exact meaning of panda is hard to specify but it usually denotes some uncertainty on the part of the speaker. kave nda ndsremi mbeli ke wundu banda naa (NY2) give it poverty one to person thus ? Should one make another poor, do you think? xwa gamba que asé tenza banda Perhaps that girl is in the in bush girl that perhaps bush, I don't know. (PSh) b.§mma This expletive is normally used to mean something like, "even so" or "never-the-less". katexwé nde sanda He will run never-the-less. will-run he never-the-less (Gr7l) dla 'ya meneté sanda I must do it even so. (Gr7l) only I do never-the-less c. wusi This is used to mean something like, "you say" or "you don't say" in English. ... gwene kata tli ass wp§i_ . . . salt to cook the meat, you salt cook meat that you-say say? (CP2) 'ya wusi Luc It's me, you don't say, Luc. I you-say Luke (CP2) construe ic and a 1n aiscc illil D" p) (I) anothe kn dSt < .2 88 2 . 5 CONSTRUCTION CLASSES . This analysis of construction classes is an attempt to posit constructs in the grammatical stratum. It considers discourses as bas- ic and defines clauses as constructions which can fill positional slots in discourses, phrases as filling positional slots in clauses and words as filling positional slots in phrases. Sentences are postulated as another order of construction, that is, as fruitful analytical con- structs but outside of the framework sketched here. They cannot be defined as simply filling a positional slot, they are particular kinds of clause combinations (cf., 2.3.5 below). The brief sketch of these major construction classes in Kapsi- ki is designed to serve as background to the more detailed discussion of verbs in the following section. Each class is briefly considered by means of a working definition, a review of major sub-types and a number of illustrations. 2.5.1 Phrases. Phrases may be defined as constructions which can fill clause slots. ’Most phrases are constructed around either verbs, forming verb ' phrases (VP), or nominals, forming nominal phrases (NP). 2.5.1.1 Nominal Phrases (NP) Nominal phrases are constructions with a nominal head (Nh) which may fill a time (T), subject (3), indirect object (IO), object (O) or satelite (sat) slot in a clause. There are three main types of-nominal phrases, distinguished by the kinds of relationships ex- hibited between dependent words and the Nh. In addition, a few unclas- sifiable nominal phrases will be noted. a. Genitival Noun Phrases (NPgen) NPgen take the shape, Nh plus genitival particle Nd. The Nh may be any noun except locatives, time words or interrogatives. The Nh determines the shape of the genitival particle as discussed in 2.2.l.lb and the dependent nominal (Nd) can be a class I noun, a per- sonal pronoun (cf., 2.2.1.2) or another NP. Examples: I I'll 89 (1) noun plus noun xwéta ta kwara donkey skin skin of donkey (2) noun plus prounoun gs a nci their house house of them (3) noun plus NPgen gena a meleme ta nci news of their village word of town of them (A) noun plus NPqual (cf., 2.3.1.1b) kweturumbé a va bak two years earlier behind of year two Examples of NPgen in clauses: (l) in a time (T) slot kweturumbé a va bak gala wumevé Two years before he had mar- behind of year two marry ied. (TS2) vatamd’al pels Garewa Last year (I) was at Garoua. last year on Garoua (TS2) (2) in a subject (S) slot gala ta Kwejimté nde fa The god of Kwejimte hears. god of Kwejimte he hear (NAB) sé yasé ye yite a da le miyi da (WSS22 come come father of me with mother of me My father and my mother are coming. (3) in a satelite (sat) slot wundu kaseka dageva yita wuie ta kwa zerema (FM6) person come discuss marriage for girl of which not loved Some one is coming to discuss marriage for the child of the unloved wife. katemené tlené geg'yé 1e wuzege ta kaxeci (TSh) will-do work we with child of Kortchi We will work with someone from Kortchi. (u) in an object (0) slot katevindiyé gena a meleme ta nci 'ya (TSH) will-write word of village of them I I will write the language of their village. lAlthough vatamda "last year" is written as one word in this thesis, it is derived from the genitive noun phrase *[va ta mbedé] ‘which would be translated literally, "year of yesterday". In fact, this latter construction (which is starred) never is heard in actual speech. 90 (5) in an indirect object (IO) slot 'a 'ya kémenté nda kc gu a da I did it for myself. ( 8510) I did it to head of me mene nda nde ke za a gks She does it for her husband. do it she to man of her (PSI) b. Qualificative Noun Phrases (NPqual) Qualificative noun phrases take the shape Nh plus qualifier or qualifier plus Nh (cf., 2.2.3 for a discussion of word order). NPqual's can fill any clause level slot filled by nominals,tmlring semological restrictions. Most NPqual's are of one of the following types: (1) ideophone plus Nh ggmégémegpelé nyi taga There are red stones there. red red stone these there (WSh60) wusi fét kwa kwetlekwa gké Everything is in her basket. things all in basket of her (P22) (2) Nh plus numeral pa gs mcéf te gkwa ass There are five houses there. houses 5 on place that (3) Nh plus demonstrative male nya kédepe. This woman knows (PMl) woman this knows ' wuzege asé va megeké That child is at his mothers. child that is mother (PMl) petleké wundu nyaga mbeli Someone killed that person. killed person that one (PMl) (h) Nh plus specifier pa kweté tasa mbe dzeve (He) has a certain plate in certain plate in hand (PJh) kweci mbeli kwa kwembewalé Some people are in a boat. some peOple in boat (PJh) c. Prepositional Noun Phrases (NPprep) Prepositional noun phrases take the shape, preposition plus Nh. The Nh may be any noun except locatives, timcwords or interroga- tives, and personal pronouns or other NP's. NPprep's ordinarily fill a satelite slot in the clause, phrases using the preposition l2 may be marked with the coordinating particle ye and serve as a coordinate subject or Object. (l) preposition plus noun 91 tlamté le xwa Cut (it) with a knife! (wshou) cut with knife 'a xwa te xedi The knife fell to the ground. knife on ground (WSMOO) (2) preposition plus pronoun ssyi kwetég wundu nya kexevé kelepé xwa xé (P21) only one person this caught fish among them Only one of them caught any fish. bak kwembewalé 22. Two boats are on it. (P03) two boat on it (3) preposition plus NP te kweté wusu In a certain thing. (PS3) on certain thing xé le za a flké 1e wale a gké (P22) they with man of her with children of her They, with her husband and with her children. (M) coordinator plus preposition plus Nh mené tlené y§_ Davité le Yakubu David and Jacob will work. do work David with Jacob (TSh) taga .XE gwezfi le Aye wufé There are trees and grass. there grass with trees (PSh) d. Miscellaneous Noun Phrases There are several constructions, sometimes treated as single words, which appear to be NP's. Even though the Nh may be identifiable in these constructions, the type of relationship expressed between it and the dependent words is unique to the construction and thus cannot be classified. For example: wusi ge seda animals (GrhG) things foot - wundu 9e 18 human being (Grho) person ? ? wusu ke zemé food (WSQOS) thing eat There are also several constructions in general use employing the relative kwa'but which are not relative clauses. For example: kwa gwelé left (as in left hand) (P02) kwa zerema the unloved one (FM6) kwa juni the loved one (FM6) and I‘ll verb ha: lm' IH In {It $1.33 3.1 ) 9 ‘ . m5 “.1” Us t5 .- S ~ CC S ~ C $501»: ‘6, n 92 2.3.1.2 Verb Phrases (VP). Verb phrases are constructions built around a verb head (Vh) and filling a slot in a clause. There are three main types of V15 verb marker (Vm) plus Vh, dependent verb (Vd) plus Vh and adverb (adv) plus Vh. These ordinarily function to fill the predicate slot in the clause. In addition there are a number of locutions built around verbs which, like the miscellaneous NP's, are difficult to classify. (l) Vm plus Vh ‘pa 'ya dzaa Teki Then I went to Teki. (T82) then I go Teki 'a kémpavé (He) hit (him). (WSSBl) hit (2) Vd plus Vh késé 'ya g§_ wé ... I didn't want . . . . (SS9) come I wish not 'a késé tsa lewela (He) cried out. (WS301) come cut cry 'a sé zeme dafa wé (He) isn't eating. (SSlO) able eat mush not (3) adv plus Vh kwa'a ketsaks ga Pow: It strikes you. (NAl) pow cut you ntigu kasate 1e Suddenly (he) came with (it). suddenly come with (FM6) nzaké kéleté téréké beteyi te ... (PMl) pow took splash poured on (She) grabbed it and 'splash' (she) poured it on . . . . Miscellaneous Verbal Constructions The verb sé "come" and dag "go" with various extension suf- fixes, are frequently found in the type of phrase, verb plus nya plus verb with the meaning, "it came to pass" or "it happened". The ex- tension suffix, when used, indicates how it came about. For example: sé nya sé wundu asé ... It happened that the person come which come person that referred to . . . . (FCY) sate nya sate nde ... It happened that he . . . . come which come he (FC?) In addition, several of the conjunctions and other function words, although treated as single units since their constituent parts can not be easily identified, appear to be constructed around verb 93 roots. For example: vagyanza "because" appears to be composed 0f.X§ plus the de- monstrative, nya and the verb stem nza "stay", "exist". 2.3.2 Clauses. Clauses have been defined as constructions which fill positions in discourses. Functionally, however, they reflect the semology and can be defined as the grammatical realizations of event configurations. Reflecting this theoretical orientation, clauses in Kapsiki may be sub- typed as Verbal and Non-verbal. Verbal clauses are the realizations of configurations in which the event is conceived as an action and non-verbal clauses the realizations of event configurations with enti- ties, relationals or descriptions as events. (The terminology is based on that of Gleason, 1968 and Stennes, 1969.) 2.3.2.1 Verbal Clauses. a. Clause Slots Verbal clauses contain an obligatory predicate (PRD) slot and optional subject (SUB), object (OBJ), indirect object (IOB), locative (LOC), temporal(TMP), modifier (MOD) and satelite (SAT) slots. The fillers of SUB, OBJ and IOB slots are generally the realizations of concomitants playing participant roles in the semology while LOC, TMP, MOD and SAT slot fillers are the realizations of circumstantials. The presence or absence of the first three slots are used as the defining characteristics of the major clause sub-types. In addition to being generally filled by the realizations of participant sememes, these slots stand positionally in the closest proximity to the PRD slot and with some verbs are obligatorily filled. The order of slots in the clause is not rigidly set in Kapsiki; however, some general tendencies can be noted. Certain forms of the verb usually precede the subject while certain others ordinarily fol- low it. The present progressive form of the verb is always formed by placing the OBJ before the verb and the SUB after. Furthermore, in keeping with what has been noted concerning word order of nouns and noun modifiers (ideophones) the clause can be considered to be verb- centric. That.is, the primary slots are found in closest proximity l-‘c' .Ot' ... J found “ s. P . 91* to the verb, either following or preceding it and secondary slots are found farthest removed. For example: gala xs I klsketi I kambsxwé I va degwava uwé (SP1) all night dogs bark because hyena TMP SUB PRD SAT IsecondaryI primary I secondary The dogs barked all night because of the hyenas. katexi I gs a da I 'ya I xénde I 1e pélé (SP1) will-fix house my I tomorrow with stones PRD OBJ SUB TMP SAT I primary I secondary Tomorrow, I will repair my house with stones. kwa gwa I sé I male nya I le simfi I yita dzeve ta gks in river come woman this with broom beside hand of her LOC PRD SUB SAT SAT Isecon. I primary I secondary This woman came from the river with a broom in her hand. (PMl) The following brief discussion of the clause slots should clarify the examples of clause sub-types. (1) PRD slot The PRD slot is obligatory in a verbal clause. It may be fill- ed by a verb or a VP. If the filler is a VP it may take any of the shapes described above (cf., 2.3.1.2) or a Vh morphologically reali- zed as /¢/ plus an adverb (cf., 2.2.3.1a). The form of the PRD filler determines the order and the obligatoriness of the other primary slots. For example, if the PRD is filled by a VP the SUB falls between the dependent word or phrase and the Vh. If the verb is in the progres- sive form the OBJ will precede the verb and the SUB follow it. Other- wise the normal order is, PRD (IOB) OBJ SUB. This is by no means fixed and much stylistic variation occurs. (2) SUB slot The filler of the SUB can be either a nominal or a NP. If it is a pronoun third person singular it usually takes the phonological form /¢/. In most cases, although not necessarily, the SUB filler is the realization of an entity sememe. (s) OBJ slot Like the SUB this slot is ordinarily filled with a nominal or NP. Many verbs obligatorily take an object, however; the common col— location of PRD-OBJ in the surface grammar is the realization of a a manner ‘ 4‘5 tne uni... in. ‘he sets J. L ... ‘n~‘o WI .. _ T°L12€Q eater var ’fi.‘ . ’14 '2 I LU.) a ‘1 ‘1- 3:)“ r: ~- mo... ¢.\_ 95 a number of very different relationships at a deeper level. That is, the units filling the OBJ slot are the realizations of sememes which stand.in a number of different relationships to the event sememes in the semology. These differences are dictated by semological categories into which the events fall. For example, verbs of motion often require a locative sememe which is connected to it by a goal relational and is realized grammatically as an object. This type of object is rare with other verbs. (h) IOB slot This slot is filled by either a personal pronoun of the first or second person oblique case, or by'gdat'When filled by nda, if the indirect object is a third person singular personal pronoun form.§da stands by itself. If the indirect object is a third person plural pro- noun or a noun, the pronoun or noun is placed in a SAT slot as a NPprep of the type knglus Nh. Examples of the ICE slot follow. 'a kémeneté I.§é I tlené (He) worked for me (WSlh) did me ‘work PRD IOB OBJ 'a kémeneté I nda I tlené I ke xé (He) worked for them. (WSlh) did it work to they PRD IOB OBJ SAT geze I nda Ike yite a ya I sé nde Tell your father to comet say it to father you come he (WS331) PRD IOB SAT OBJ (5) TMP slot This slot is usually filled by a time noun or a NP which is the realization of a circumstantial serving to specify the temporal setting of the event. The slot is optional with all types of event realizations. Unless preceded or followed by a MOD slot, it normally comes either clause initially or finally. (5) LOC slot This slot is normally filled by either a time noun or a HP which is the realization of a circumstantial serving to specify the spatial setting of the event. It may either precede or follow the primary slot configuration. This slot is optional with all classes of verbs. With verbs of motion it is replaced by an OBJ slot. ()0 (7) MOD slot This slot is almost exclusively filled by ideophones serving to qualify the action of the verb or the clause as a whole. It is option- ally filled in all types of phrases. For example: 'a I 'ya I képamté I maruré I kwélékwelé I sold lots of rice. sold rice much (WS372) SUB PRD OBJ MOD wa nde I geze I ya I waa Who told you? (WSSl2) who say you ? SUB PRD IOB MOD (8) SAT slot The SAT slot is filled with constructions expressing a variety of relationships to the clause core. Included are NPprep's indicating an instrumental or accompaniment relationship of a circumstantial to the main event, verbs, VP's or dependent clauses realizing events which stand in a purposive relationship to the main event in the semology, and the realization of several other types of circumstantials. (Cf., for example, the discussion of the IOB slot fillers above.) b. Clause Sub-Typgs Since the only obligatory slot in verbal clauses is the PRD, verbal clauses can be represented as having the canonical shape: PRD (IOB) (OBJ) (SUB)l Only the primary slots, that is, those normally filled by-the realiza- tions of participant sememes are represented since the secondary slots may be optionally filled in any of the verbal clause sub-types and are not diagnostic. This cannonical form yields eight possible sub-types. A few of the possible types, namely, the PRD OBJ and PRD IOB OBJ con- figurations are relatively rare in my corpus. They regularly occur with the realization of only a very few event class s. (l) PRD only This clause type only occurs with verbs in the imperative. IzeméI Eat: (Cth) PRD Izexwfi I cfikI Eat everything: (Cru8) eat all PRD MOD 1Items in parentheses are optional. 97 (2) PRD IOB Igeze I nda I biyitsaI say him again PRD IOB MOD Igeze I daI say me PRD IOB (ws196) Tell him again! Tell me! (SS7) This is the only verb used with this clause type in my corpus. I It could be analyzed as exhibiting an OBJ slot filled with a morpheme of /¢/ phonemic shape. (so PRD OBJ Ikélawiyi I lekwesa nyaga I tagaI there PRD OBJ LOC put-up cloth that Ipeliyi I kesi I te xsdiI put mat on ground PRD OBJ LOC (A) PRD SUB Ima I kwansexevé I geg'yéI arrived we MOD PRD SUB I'a I tli nya I kédeI meat this spoil PRD SUB . PRD I'a I I ké 'wusiI he laugh PRD SUB PRD (5) PRD IOB OBJ Igeze I nda I ke xé I sé xéI say to them come they PRD IOB SAT OBJ Inta I da I 'wfi'qu give me money PRD IOB OBJ (é) PRD OBJ SUB Ipa I wuzé a miyi da I késevé I ksméI then son of mother my caught PRD SUB PRD I'ya I temené I tlené I besiI I will-do work today SUB PRD OBJ LOC (j) PRD IOB SUB Hang the cloth there! (WSMOQ) Spread the mat on the ground. (wsuoo) After we arrived . . . . (TS2) This meat is spoiled. (He) laughed. (ws2u5) Tell them to come. (wssse) Give me some money. (WS3l8) My brother caught a duiker duiker. (TS? OBJ I will work today. (TS?) 98 Imene I nda Inde IkeEa za a gksI She does like this for her husband. do she for man of her (P28) PRD IOB SUB SAT Iwune a gké I kais I ndaI He is unhappy. (PSI) heart of him pains to-him SUB PRD IOB (a) PRD IOB OBJ SUB Itlené I yamene I da I ndeI He is working for me. (WSlh) work ing-do ~me he OBJ PRD IOB SUB I'a I 'ya I kégwenegfi I ndaI'wfi'qu He sent money to him.(WSS98) I sent to-him money PRD SUB PRD IOB OBJ 2.3.2.2 Non-verbal Clauses . Non-verbal clauses can be conceptualized as clauses built around the realizations of event sememes which are not actions. The most com- mon type of non-verbal clauses are descriptive. Reflecting the semology of which they are realizations, descriptive clauses may be classified as: (a) locational, (b) equational and (c) qualificative. Non-verbal descriptive clauses are, for the most part, the re- alizations of event configurations built around a descriptive event se- meme. However, they are only one of the possible grammatical realiza- tions of descriptive event configurations. For example, locational descriptions are often realized as a V or a V5 (cf., 2.2.2.2, 2.2.2.5) 2 and qualificative descriptions as a V in a non-active form. Conversely, identical grammatical constructions miy be the realizations of very different relationships in the semological stratum. (Cf., for example, the OBJ SUB type equational clauses below, where in one case the par- ticipant-event relationship exhibits a possession valence and the other a description valence.) Non-verbal clauses may be analyzed as having an obligatorily filled SUB slot. However, this is not equivalent to the SUB slot in a verbal clause which is normally filled by the realization of a partici- pant standing in an agentive relationship to the action event in the semological stratum. Neither is the SUB slot filler necessarily the realization of the event. The event may be an entity or a relational realized as a LOC, MOD or OBJ filler. The SUB will reflect one of 99 several possible relationships to the event, such as, possesion, equa- tion, location, recipient or interrogation. a. Locational Clauses These are normally realizations of event configurations con- structed around an event realized as a LOC slot filler, that is a N? or a NPprep. Ite gwu I dugqu The pot is on the fire. (WShQL) on fire pot LOC SUB Ikwembewele I teke | mbe Kapsuku | in Are there canoes in the canoes here in Kapsiki ? Kapsiki country? (PMl) SUB LOC LOC MOD Itékwatakwa I wuve a da I vare nagkéI My farm is near his. near-near farm of me by his MOD SUB LOC b. Equational Clauses Equational clauses are composed of an OBJ and a SUB slot both of which are usually the realizations of entities which are related by possession, description or interrogation valences. Itsema I wuzege ta gal How is your child? (CPl) how child of you OBJ SUB Igada I wuzegeteza nyaI This boy is my son. (WS356) mine child-man this OBJ SUB c. Qualificative Clauses Qualificative clauses are normally composed of a MOD slot and a SUB slot, the filler of the MOD being a qualifier or a non-active form of a V . This filler is the realization of a descriptive event I sememe in the semology. Idibédibé I kwe nyaga] That goat is fat. (WSu33) fat-fat goat that MOD SUB Imcélfi I ndeI He is rich. (WSh88) rich he MOD SUB Imégéla I nyaI This is good. (WSHBI) good this MOD SUB 100 2.3.2.3 ya Clauses- The event-participant or event-circumstantial relationship which is normally not formally marked grammatically in the reaization of de- scriptive event configurations (cf., the examples above in 2.5.3.?) is sometimes realized grammatically as yg. This is possibly the same mor- pheme, having the same phonological shape, used to relate an object to a verb in the continuous form of verbs. However, thkstype of clause, illustrated below, is very rare in my corpus. Isisi I ya Imbe dleraI The snake is under the rock. snake is under rock (GrlS) SUB LOC Iwusu kévindi I.Xé.I te tabeléI The pen is on the table. thing to-write is on table (GrlS) SUB LOC Compare the above with the following example of y§_as the OBJ- PRD relator (analyzed as a verb prefix) in the continuous form of the verb. Imé Iwé Ival fall falling rain OBJ PRD SUB The rain is falling. (GrlS) 2.3.3 Sentences. Sentences may be defined as concatenations of clauses exhibit- ing a formally marked grammatical relationship. As such they consti- tute a different order of construction from phrases, clauses and dis- courses in that they don't fill a positional slot in a higher lever construction. The grammatical relationship exhibited between the clauses of 8- sentence may take one of the following forms: (1) temporal,(2)depen- dent or subordinate, and (3) logical. 2 .3.3.1 Temporally Linked Clauses . Temporal linkage of clauses is ordinarily marked by verb mark- fires or affixes which are the realizations of sequentials in the seme- Imic: structure. (These are treated in detail in section three and will Orthy be briefly illustrated below.) Clauses may be linked to show simple succession of action, “ulétt one action was finished before another started, that two actions lOl take place simultaneously, that a state was interrupted by an action or that one action was started while another was in progress. For example: a. Simple Succession I23 I wuze a miyi da I késevé I kemé ”‘22 I xkaké I g'yéIl then son or mother me caught duiker then called us mkr SUB . PRD OBJ mkr PRD OBJ My brother killed a duiker and then called us. (TS2) b. One Action Completed Before Another Ima I kwansexevé I geg'yé H ntaké I geg'yé I kawutséI (TS2) after-came we started we dance MOD mkr PRD SUB MOD SUB PRD c. One Action Interrupting Another Ima I nda I naxe I tlené H naxé Inde I kédzegwa I géI (Gr62) before finish work finish he go himself MOD mkr PRD OBJ PRD SUB PRD OBJ d. An Action Interrupting a State Imetleté I geg'yé H nda I dzembé I ndeI We were standing when he stand we when go he went in. (GrSS) mkr PRD SUB mkr PRD SUB 2.3.3.2 subordinate Linked Clauses. Sentences composed of subordinate linked clauses are of three types, reflecting different kinds of dependent relationships between actions or states. These are: (a) relative clauses, (b) purposive clauses, and (c) auxiliary clauses. a. Sentences with Relative Dependent Clauses Iwundfi. H BEE. I geze I nya I 'ya I géni H naxé késéI person which say which I then finish come SUB rel PRD rel SUBd MOD mkr PRD b. Sentences with Purposive Dependent Clauses These sentences are marked in two ways, possibly realizing (iifferent, but hard to specify, relationships on the semological stra- tlnn. One type is marked with the prefix /k§-~ £27; The second is Inarked.with the marker (mkr) keZa preceding the verb in the purposiVe ‘ . 1In these examples phrases are seperated by a single vertical 1line, (I) and clauses by double vertical lines, (H). Dependent Clifluse slots are marked with a small 'd', for example, PRDd. 102 clause. These represent only two of three possible realizations of a single relationship between event configurations on the semological» stratum The third possible type of grammatical realization, is by the use of a sentence composed of logically related clauses (cf., 2.3.3.3). Ipa xkaké I 'y’ u keza zemeI Then he called us to eat. then call us for eat (TS2) PRD OBJ mkr PRDd | pa I geD'yé I naké H kamexa I nda I gqu (TSh) then we saw light to-it fire mkr SUB PRD PRDd IOB OBJ Then we saw the judgement of . . . . Ikweci mbeli I sé H tenewe I ya I Some people came to see you. some people came to-see you (WSlB9) SUB PRD PRDd OBJ c. Sentences with Auxiliary Dependent Clauses. The expression of modality, temporality, negation and even of completion is sometimes realized on the grammatical stratum by the use of auxiliary verbs. This type of construction can be conviently ana- lyzed as a complex clausal unit, that is, a sentence. The most com- mon auxiliary verbs are: 93 "wish", §§_"come" (used to indicate past action), and dema "not". I'a I 'ya Ikégé H kaseméIl I wanted to eat. (Gr3h) I wish to-eat mkr SUB PRDa PRD Ikédema I wundu H kakelé ...I A person does not take . . . . not person take (8PM) PRDa SUB PRD I'a késé H tsa I 'yitlaI (He) coughed (WSh20) come cut cough PRDa PRD OBJ 2.3.3.3 Sentences with Logically Linked Clauses . The term 'logical linkage' is used with reference to clauses ‘Jhich may be related grammatically by: (a) a clause level conjunction, (b) a negative marker, or (c) an interrogative marker. a. Conjunctive Sentences Intigfi I geg'yé I kasake I xetwedi I| v_a_nyanza II 'ya I kamene I suddenly we came yesterday because l do MOD SUB PRD TM? conj SUB PRD 1The small 'a' in these examples means "auxiliary verb". 103 Example, continued from page 102. tlené I besi I work today We returned home yesterday because I had OBJ TMP work to do today. lee) Igegkfi I sé H keza zeme I wusu H ba'a H keZa mene I tlenéI He came for eat thing and for do work SUB PRD PRDd OBJ conj PRD OBJ He came to eat something and to do some work. (Gr73) b. Negative Complex Sentences The negative marker y§_may be used to negate a number of clauses thereby relating them logically into a sentence. Generally if several related clauses are to be negated some form of the auxiliary verb fig. "able" precedes them and the marker HE comes at the end of the entire construction. I'a St N xwa nya I kédema H kazeme H katlave I tli I wéI able knife this not to-eat to-cut meat not PRDa SUB PRDa PRDd PRD OBJ MOD This knife won’t cut the meat. (ws268) Ikédepfi I kwa nya H 11 I xwé kwa a§€ H mbeli I te I EEI know place this go road that person on not PRD OBJ PRD SUB SUB LOC MOD No one knows where this road goes. (P03) c. Interrogative Marked Sentences The interrogative markers y; and nee behave in much the same ~way as the negative marker, that is, they can serve to make several preceding clauses interrogative and thus link them into one sentence. Ipfikwfi I teli I gké H depfi I wundu I naaI (P03) far goes it know person ? MOD PRD SUB PRD SUB MOD 2.3.h Discourses . Discourses constitute "the true minimal free form" (Hoenigswald, 1960. l) of the language. A discourse may comprise a single phrase, or it may, theoreticallx contain an infinite number of clauses. ‘Some discourses, such as fables and conversations, may have a formally marked beginning and ending. Others while lacking these features, reflect a logical relatedness of the constituent clauses. This is not to be con- fused with the logically marked linkage forming the sentence types dis- cussed in 2.3.3.3. Discourses are marked grammatically by; (a) the use of proforms, th (b) optionally being enclosed in the formal openers and closers refer- red to above, and (e) by being preceded and followed by silence. Reflecting the semological structure of the language, discourses can be conceived of as the grammatical realization of complexes of event configurations. Following the lead provided by Taber and Stennes, who have published extensive semological analyses of African languages (cf., Taber, 1966 and Stennes, 1969.), it seems fruitful to conceptualize Ka- psiki discourse as being constructed around an event line. (In fact, this concept has underlain the discussion of the grammar up to this point.) The event line is composed of a series of events realized gram- matically as predicates, if they are action events, or in various other ways if they are not. Related to the events and to each other in a number of different ways (referred to as valences) may be other sememic units, that is, par- ticipants and circumstantials. The set of participants, circumstantials and event, along with the set of relationships exhibited among the parts, constitutes an event configuration. This is realized grammatically as a clause. The total complex of related event configurations is realiz- ed grammatically as a discourse and any break in the event line will be realized in the grammar as a discourse boundary. No attempt is made here to devise a formal taxonomy of discourse types. However, among those used in the corpus under consideration for this thesis are (l) fables (gena xéca "stories of old"), (2) sermons, (3) conversations, (h) narratives (general accounts of events or des- criptions), and (5) addresses. The several examples of formal openers and closers are followed by an example of a discourse analyzed to pro- vide illustration of all the construction classes discussed in section 2.3. 2.3.h.l Discourse Types. a. Fables Fables are usually introduced and concluded formally. They are recited frequently by the old men of the village after the evening meal in an area set apart for gatherings in each section of the village. They are also frequently cited during 'judgements' to provide some type of precedent and are generally known by everyone in the village. 105 The commonest discourse opener is pekwuké (meaning uncertain) either by itself or in a phrase. For example: pekwuké pa kweté za ... Here is a story. A certain then certain man man . . . . (FM6) pekwuké késé nde Here is the story of how he came he came. (FM3) pekwuké kweté Here is another story (FML) another At times the formal opener is dispensed with and the fable is simply introduced as a "story of old". tetla bumfi gena xéca 'ya gé I will tell us a fable then. will-cut us word old I then (FMl) ma sé ta kaggwedu né Sunu né This is the coming of the people coming of Mbgodians then Sunu of Mogode then, Sunu. (F07) Fables may be formally closed by the use of an ideophone not heard in any other context. The first sound is formed by pinching the nostrils with the thumb and forefinger and suddenly letting go so that the air comes out accompanied by voicing. . 855 kévawfi It is finished. (FMR) €88 kévawfi.gena xéca The story is finished. (FM3) Occasionally the ending is simply a statement to the effect that this was a fable. tsaga kweté gena xéca nya vesi This was another old story of there another story this before before. (FMl) wawa ké mbeli gegké tsetsé sé nde Thus say people, his coming thus say people he now come he was like this. (FC7) b. Conversations Conversations are introduced in various ways, usually by some reference to the time of day and how the person engaged has spent the preceding segment of it. For example: 'a na kévegwfi Luc? You have spent the day Luc? you spend-day Luc (CP6) 'a na kéxe nee Sunfi You have spent the night, Sunfi? you sleep ? Sunfi Conversations often end with a reference to a future meeting. (Pierre) scyi xési Until later. (CP3) only a little while (Luc) yéwfi O. K. O. K. 106 c. general Most discourses have no formal opener or closer. However, a general topic may be introduced by explicitly calling attention to it. For example: wa nya Zera nya ... Here then is Zera . . . . (CD9) here Zera this Any discourse can be introduced with the introducer mg, the verb marker pa or the expletive of attention, yéwfi "O.K.". For ex- ample: ;mé gena nya nde nya n1 This then is the story. (Nh) word this it this then pawaa mbe dzeve ta gké He has a knife in' his hand. then knife in hand.of him. I (PMl) yéwfi pa sala ndeke g'yé wusegé O.K., then God gave me a child. O.K. god give us child yfiwu fu mu ndedeku O.K., hear me well! (N13) O.K. hear well 2.3.h.2 Discouse Analyzed. The following short narrative discourse is a description of a picture. Words are separated by space, phrases by a single vertical line (I), phrases embedded other phrases are marked by brackets ([...]) clauses are enclosed in double vertical lines (H) and sentences by parentheses (...). Below each word is a word by word gloss, below this the word class is indicated and finally phrase types and clause slots are marked. Slots are indicated to the right of the hyphen and the filler to the left. Thus, NRgual-SUB would read, "subject slot filled.with a qualifying noun phrase." The clauses are numbered con- secutively by an arabie superscript to their left. lIIkweté nya wundu I kaya I wusi a gké II (2 II kad‘é II:5 nde I certain this person guarding thing of him wants he spc dem Nl Vl Nl pro Vl pro NPqual-SUB PRD NPgen-OBJ PRDa SUB kali H hkatewumé I malé H) 5II[ 86 a nci ] nyigi I dziyi I taga I go to-marry woman house of them those go there V2 V1 N1 N1 pro dem V2 N2 PRD PRDd OBJ NPgen-Nh of NPqual PRD LOC 107 6 le yt [ xa ta nci ] H xawanya I nya I nee H 7naxe I kwe ta with and millet of them corn this is it finish goat of prep coor Nl pro Nl dem lkr Vl Nl NPprep with NPgen for Nh-SAT OBJ SUB MOD PRDd NPgen- gké I kémbele Igé H 8 kaxwé I nde H 9béréda béréda béréda I him escaped self running it gallop, gallop, gallop pro Vl pro Vl pro adv SUB Vh of VP-PRD OBJ BED SUB MOD 1 ¢ I kacadb H ( 0” kateli 11 H katekesé I ¢ I ya 12 H kweté I it running will-go will-catch it if not or other pro V1 V2 V1 pro conj spc SUB Vh of VP-PRED PRD PRDd OBJ OBJ kali 13 H katewumé I male ta gké I le H) go will-marry woman of him with V2 V1 Nl pro prep PRD PRDd NPgen-OBJ NPprep-SAT Translation This other man is guarding this thing. He wants to go and marry a woman (with it). That is their house over there along with his millet. Is that corn, instead? The goat has escaped. It is running. Clippity, clippity, clippity it is running away. He will go and catch it or if not he will catch another to marry his wife with. ‘ (PM6) The above discourse serves to illustrate the following clause types: Verbal independent clauses: l, 3, S, 7, 8, 9, 10, 12. Verbal dependent auxiliary: 2. Verbal dependent purposive: h, ll, 13. Non-verbal equational: 6. Discourse markers The agent participant, CERTAIN PERSON, is realized as a Nl-SUB when it is introduced in the first clause. Thereafter a pronoun is used. The other major participant, GOAT, is realized when it is intro- duced in clause 7 as a N serving as the Nh of a NPgen and whenever 1 after referred to is realized as a pronoun. There is no formal Opener or closer, as is normal with narrative discourses. SECTION THREE VERBS 3.0 INTRODUCTION This section is a detailed summary of the grammatical sketch which has preceded. It is an attempt to analyze the verbs as gram- matical units in their grammatical context. However, this is done in a further attempt to specify the semological realizates of the verbal constructions as a step in gaining insight into the Kapsiki's cul- turally determined conception of "reality", hence the relevance of the ethnographic sketch constituting the first section of the thesis. Some further clarification of the theoretical construct sup- porting this analysis is in order at this point. As already indicated (section 2.0) the analysis so far has been presented within the general framework of stratificational theory. That is, the language is viewed as essentially a coding device capable of transducing information con- cerning the world of reality as the Kapsiki perceives it, into a par- ticular set of vocal symbols. This mechanism is conceived of as con- taining at least three strata, the phonological, the grammatical and the semological. The focus has been placed on the explanation of the grammatical stratum. The rationale for this emphasis is the contention that an analysis which specifies the grammatical units (morphemes and lexemes) and their arrangement (tactics or syntax) into symbols will permit the analysis of sememes which in turn have an isomorphic relationship to the Kapsiki's perception of reality. With this theoretical orientation, the verbs will be examined from the standpoint of the grammatical forms they exhibit. The struc- turing of these grammatical units provides the main clue as to how the Kapsiki have learned to structure the world of experience about which they talk. Since verbs are normally the realizations of action events, the various grammatical forms should indicate Kapsiki belief concerning the nature of action. Focusing on the overt grammatical markings which 108 109 convey information about actions, several types of semological features will be investigated. These include the expression of modality, tem- porality, aspect, roles and semantic field as segmented and classified. Implicit iJI this approach is the assumption that actions have some kind of "physical reality" Which exists apart from the abstract classifications imposed on them by a perceiving human being. (Obviously to talk of segmenting bits of reality units called actions is a tacit admission that one is already operating on a certain level of abstrac- tion. This is analagous to making the statement that phones are sounds and units of speech and not language, because language is an abstraction and speech is its physical realization. In fact, sounds themselves (phones) are more or less arbitarily selected segments of a continuum and thus represent a certain level of abstraction.) This unavoidable disclaimer notwithstanding, just as it proves fruitful to distinguish between phones and phonemes for analytical pur- poses, it seems useful to proceed on the assumption that actions can conceptually-~if not empirically--be described as existing apart from the structure of any expression system into which information about it is transduced. From this point of view, any action can be conceived as (1) having existence in time, (2) having existence in space, (3) being re- lated in some way to the one speaking about it, (t) exhibiting certain characteristics relative to itself, and (5) enjoying a relationship relative to participants. In addition, (6) any action is itself a symbolizable part of "reality" differing, in various culturally de- termined degrees, from other parts. Quite obviously there may be other dimensions which the writer's own culture either precludes his conceiv- ing or at best makes conceptualization very difficult. Although these dimensions of action can in some sense be con- sidered universals—-maybe in the sense that pitch and intensity are sound universals--the manners in which they are conceived, chosen for expression and expressed, are unique to any given language. For example, it is conceivable that the point in time when an action takes place is not considered important every time the action is referred to so that this is not marked obligatorily in the grammar. Or, if time is marked, it can be viewed as relative to some point other than the speaker's 110 present. (In fact this seems to be the case in Kapsiki where verbs are not ordinarily marked to show tense but sequentials are commonly real- ized grammatically reflecting when one action took place relative to another action or a described state.) In Kapsiki, as expected, no single set of grammatical features related to verbs, corresponds exactly to the possible dimensions of action sketched in the paragraphs above. As each feature is examined, however, an attempt will be made to indicate its role (non-technical sense) in expressing the Kapsiki's conception of these dimensions of action. 3.1 GRAMMATICAL FEATURES This section reviews the main grammatical features realized in verbal constructions, including (1) the root, (2) the stem and (3) markers and prefixes. The following section will discuss the semological features realized by these grammatical features. 3.1.1 Verb Roots. All verbal constructions have as a minimal form a verb root. For example, the imperative §é_ "Comei", although it may constitute an entire discourse, is simply a verb root. The normal phonological shape of verb roots has been reviewed in 2.1. A verb root may be defined as the morphologically simplest grammatical realization of an action event sememe. This is not to say that verb roots are the realizations of what are in some sense simpler sememes than are stems. In fact, quite the contrast may be true. A Kapsiki verb root may be the realization of a sememe refering to a class of actions or, at least, a generalized conception of action, where- as the use of a stem, indicates that the action can be considered more particularizedl or restricted in scope. 1The notion of general and particular is basic to an understand- ing of event semology and the reflection of this in the grammar is what led.Hoffman to contend that Kapsiki verb roots are basically imperfec- tive while stems are perfective. In fact, this is not strictly ac- curate even though there is a statistical correlation between the use of extended roots and perfective markers. The reason for this is simply that stems are the realizations of particularized events and they can most easily be so specified if they have already happened (completed aspect) or are viewed as under the control of the speaker (imperative mood. 111 3.1.1.1 Distribution of Roots. An indication in the footnote below, roots are the realizations of sememes denoting a generally conceived type of action or of an on- going state. On the grammatical stratum they may occur inflected or uninflected and may fill a number of slots including, the PRD of an independentent clause or the PRD of any type of dependent clause, i.e., auxiliary, purposive or relative. Examples: sé} ta gké His coming. (Gr51) _-. come of him ...tli yadé? nde katetas ...she wants to cook meat (PMI) meat wishes she will cook a ...male nya kédepéh newéaza gké ...this woman knows how to woman this know see man her look after her husband (PMI) gaé Drink: (GRhB) 'a 922$? tlene He must work do work pééég nya kékesevé This one has finished catching (PMl) finish .this caught In the above examples the following slot filler and inflectional statuses were illustrated. (Numbers refer to superscripts fround after the root in the examples.) 1 uninflected, verbal noun 2 inflected, auxiliary PRD 3, h inflected, independent PRD 5 uninflected, dependent PRD 6 uninflected, independent PRD (imperative) 7 uninflected, independent PRD (optative) 8 uninflected, auxiliary PRD 3.1.1.2 Reduplicated Verb Roots. Kapsiki verb roots occur both completely and partially redup- licated. Although relatively rare in my corpus, the partially redup- 1icated roots are found with both the kg: and kg; inflectional suffixes when in the simple root form. Stems are also used with reduplicated roots. When in the unextended form, reduplicated roots usually denote habitual action, when used with stems they seem to indicate that the 112 action was done several times or was directed toward several goals.' For example: kamemené tlene nde He works all the time. (Gr27) do-do work he kadzedzegwa nde He walks all of the time. (Gr27) go-go walk he 'a képepesé It has been growing (at least some of it) (Gr27) did grow-grow 'a kézezemaké He has eaten(a little of several different things). did eat-eat (Gr27) . Complete reduplication is used to show emphasis. For example: I ...gwezé ...kepesépesé ...the grass...has realy grown. (P02) , grass grow-grow If the verb is used to express a noun quality, the emphatic form (with the extension suffix-mte on the first occurence of the verb root) is used to indicate that the quality exists in excess. 'a lekwesa nya kétliyaxemté tliyaxq This cloth is too narrow. did cloth this narrow-all narrow (WShS2) zewe nya kébatlimté batli This rope is too short. (Grlh) rope this short-all short 3.1.2 Verb Stems. All verb roots which occur in an active form can occur with a number of derivational extension suffixes.2 The exact meaning of a few of these extension suffixes is difficult to specify from my data. However, most of them can be shown to reflect the dimensions into which the action realized.by the root is conceived to be particularized. The extension suffixes fall into three major classes. IAlthough these "meanings" have been checked with an inform- anig the paucity of examples in my data, suggests they be considered 'vary tentative. 2In this discussion the term root will mean the unextended root and stem will mean the extended root, although technically roots can be considered stems also, in that they do occur with inflection- affixes o 3In a few cases the suffix appears to function only as a grams matical marker, indicating that the action has been completed. How- ever, this does not nullify the contention that the use of roots and stems reflects general and particularized conceptions of action. The unspecifiable cases may be due to insufficient data or may mean that the action is particularized as finished or past. (cf., Hoffman. 1966). 113 (a) those which may be used with any V1, (b) those which are normally only used with verbs of "putting" and "taking" and (c) those used normally only with'verbs of motion, that is, V2. The main extension suffixes are discussed below. a. -xwu "to do for oneself, to do all of something for the deer". 'a kévenexwu tévenexwu He vomited. did vomit-all vomit l 'a gwu kétsexwu wu§i ya da pét The fire burned all of my a“ fire burned things of me all things. (WS225) This suffix is ordinarily used with the particularizedl form I of the verbs "to eat" §3 and £2 "drink". I“ pa x6 dexwu dafa asé... Then they ate that mush...(PZ3) then they ate mush that pa geg'yfl saxwu Then we drank. (TSl) then we drank *‘a 'ya kélédcwu:5 I dug it for myself. (Grhl) did I dug h b. -gu "to do something and go away with it, or do something at one place when the effect will be felt elsewhere". 'a 'ya kégwenegEInda 'wu'we I sent money to him. (WS396) *tlegg le xwa Cut it with the knife and take it away! 1When the term 'particularized' is used to describe grammatical ftunis it refers to stems in any type of grammatical constructions where they are used--comp1etive aspect, optative and imperative moods, etc. . 2Two common stems are used to realized the semene EAT. zeme is used as the generalized form or the root and de (always with :xwu) in.t£he particularized. However, occasionally the root zeme will_ be used with the suffix -xwu in what would expectedly‘be a generalized verb. Thus: pa kazexwu:gagk£ "Then she eats hers." (P52) 3Examples preceded by an asterix (*) in this section indicate that the utterance recorded were not taken from discourses but represent examples ellicited in isolation. Due to obvious semological restric- tions, many of the possible verb forms with extension suffixes do ncfl: occur frequently. For this reason a number of 'paradigms' were elicited, then.checked for accuracy when possible with several informants. "There appears to be a semantic relationship between _g_ and -ke, the locative nominalizing suffix described in 2. 2. , and these two verb suffixes. However, no explanation for the phonological differences between _£L_and :52 is apparent from my data. 11% Obviously the uses of the extension suffixes will vary with the individual verbs. With the verb ntlé "part, divide(?)" plus the object balé "jump", in which the construction means roughly "jump" the use of :52 indicates that the action was over something, while other suffixes, -te, for example, indicated that it is particularized in other ways. Thus: 'a kéntlégg'balé dzaa gena magkwelage He jumped over the log. divide jump go over log (WS293) 'a kentletg balé jivé te xwégkwa He jumped across the road. divide jump go on road (WS29h) I J 'a kentlakawa balé pele wufé He jumped down from the tree. “‘ divide jump upon tree (WS295) c. -kel "to do something and bring here." 7 pa nde gukg'g'yt ytmu Then she dipped and brought us water. I*‘ she dip us water (TSl) *‘a we kélake I He dug and brought it. (Grhl) I dug d. -mte -mte :/mte~nte/, with the two allomorphs occuring in free variation. "the action is finished (emphatically)". pa kélemté wusu... She took off the thing... (P82) take thing 'a 'ya képampe? marure kwélékwelé I sold a lot of rice. I sell rice much ' (WS322) 'a kéntlemté zewe The rope broke. (WS375) divide rope e. -ve "to get some benefit from the action". pays g3 wusu Buy me the thing..: (PJS) buy I thing *‘a kémenexé tlene He did the work for wages. (Gr90) do work ...wundu nay kéxezé’kelepé xwa x6 ...this person among them person this catch fish among they caught fish. (PZl) if. -yi (The exact phonemic shape of this suffix is difficult to specify. The final vowel of the verb is always changed to /i/ when this suffix is used.) "to do something and leave it, not to take it up again -—-at least not soon; to do something and not expect to receive direct le., footnote #h of Section 3.1.2.1. 9The verb root pg means "sell" when used with the suffix -mte and "buy" when used with the suffix -ve. 115 benefit from the action." 'a kéntifiyi te x8di He spit on the ground. (WS2Sh) spit on ground *tliyi 1e xwa Cut it with a knife! (As grass, for cut with knife example) (Gr12) This suffix issometimes used on the cognate verb object in a <2c>ntinuous aspect form. The corpus yields insufficient data to deduce er meaning for the suffix in these constructions. xiyi_ yaxu nde She is grinding. (P22) grind grind she nsiyi yanse geva nde ... She is hurrying. (PS2) hurry hurry body she g5. -ak€ "to do a little of something, to start something and not ' finish it . " 'a kédzegakc kakwadé ama... He tried to reach it but... (WS385) try touch but pa za gks tsaxakc Then her husband started to thank her. (Gr59) man her thank *‘a ke menake tlené He started to work and then quit. (Gr90) do work Several verbs use this form regularly for the particularized fVDIfllOf the verb with no demonstrable relationship to the meaning in- Ciixzated above. For example: nak€ "saw", generalized root: newé and gwenakd "sent", general root gene kénakg 'ya wé I didn't see him. (WSl99) see I not 'a 'ya kegwenak€ sa megé I sent him to the chief (W3595) I sent—— to chief 11. :kg It is difficult to find any meaning to assign to this suffix. Hcrvwever, it is usually used with verbs of cooking or preparing food (C31?.,Hoffman, 1966) in the completive aspect and other particularized f'OI'ms and is used with other verbs in the completive aspect. It is Often found in constructions where a verb is used with a third person Sillenilar object pronoun realized as suggesting that it might in fact be a delrlasalized form of the personal pronoun gk_€. pa yabéke fluSi Then she washed her things. (P02) wash thing 116 'a kép€_k_€_ dzeve He washed his hands. (WS3h8) wash hand tsema take wusu kézeme nya na was How did you cook this food? how cook thing eat this you (W850?) 'a kénéfékg ytmu He boiled water. (WS361L) boil water serekg Leave it! i. _-_t_e This is the most commonly used of the verb suffixes and seems to frequently function as a completive aspect or past time marker. xwéljkwa meneté tlené nde was When did he work? (Gr2l) when do work he t: (this example should be contrasted with: xwégkwa temené tlené nde ‘ was, "When will he work?") pa kélétg mbete ta gks Then she took her pet. (PZ3) . take pot of her I 'a kégezepg 1e berté He spoke loudly, (with strength) speak with strong (WS253) 3.1.2.2 sé and dze Stems. The verbs sé "come" and dz_e_ "go" and sometimes vé "put" take a different set of extension suffixes with different functions and differing grammatical behavior from those used with V1. Generally Speaking, these stems indicate the nature of the terrain over which the action expressed by the root has taken or is taking place. Although ié frequently occurs as a root, either inflected or uninflectedl, 9.5.? Seldom does and is replaced by I; "leave, go" in situations where a root alone is used. (_li, on the other hand, never occurs as a stem.) Unlike V1 stems, V2 stems, while still the realizations of more Particularized sememes, often occur in generalized grammatical construc- tions. The reason is to be found in the kind of particularization they realize. Since they specify the action with reference to the terrain Over which it proceeds, or the direction it takes, this can be observed in the actual performance of the action. So, for example, one may say, kasaté x6 "They are coming up." ‘ lIt should be noted that dzegwa "walk, go", while behaving like a. root , quite probably is composed of dz_e_ plus -g_w_ra although the mean- ing of the second morpheme is not specifiable. This is one of the few eJ'Ealnples of a unit which could probably be best explained by recourse 0 a lexemic stratum. 117 The main V2 extension suffixes with their meanings are: a. -awa’ /-aa~-awa/ "to come or go down precipitously, as from a ixree or a steep embankment." The allomorphs occur, insofar as I can (ietermine, in free variation, being only stylistic or idiolectial wrariations. 'a késawa He came down. (WS275) 'a kéntlakaa balé see pele wufé He jumped down from the cut jump came down on tree tree. (WS295) 'a kweci mbeli kédzawa kadzemté... Some people went down some people go down to go to dead a to the bush. (PZl) ...nde kadzaa gwa ...she goes down to the river (P23) she go down river b. -IaIme "to go or come up precipitously, (the converse of -awa)" ntigu kasamé Suddenly she comes up (from the river) left come up (P23) same da yemu kwéline Bring me some cold water. (WXh3h) come up for me water cold 'a kédzeme wufé (He) climbed the tree. (W8272) nya male dzemé va 08... This woman is standing near her this woman come up near house house. (P02) 0. -ate "to come or go up but not too steeply" (This is used with Sé to indicate the action of standing up, but to describe the state of Ertanding normally dze plus -me, cf. the last example above, is used.) Pa 'ya sate Then I got up. (TSl) I came up ntigu 'ya kasaté... Then I left to come up... ,(TSl) abruptly I come up j kadzaté kwa The goat is going on the road up by their goes up goat house. (TSl)' xwégkwa kadzaté g8 nci road goes up house them pa 'ya dzaté g8 Lewa Then I went up to Lewa's house. (R§+) I go up house Lewa d. -iyi "to go or come down but not too steeply, (the converse of _ate)" 1In all of the examples to follow, as well as those in 3.1.2.3, When the suffix starts with /a/ the phonological changes in the final Vowel of the root are the same as those described in 2.1.3.3. . 'fl Alf”... f- 118 kasiyi nde taga kwa xwégkwa She is coming up the road come up she that in road there. (PS2) sivi teké siyi 98 nde He passes by here. (WS379) come up here come self he male kadziyi g8 a nci WOman is going to their house. woman goes up house of her (P21) dziyi 'ya vaké... I am going by here... (P21) go I by-here e. :33 (with the morphophonemic changes, the initial vowel of the Iwaot is palatalized and the vowels change. /e/ in fig becomes /a/ and fins dgg'becomes /i/ i.e., se[;ve]”§a“, and dze{_ve]“ji“ "to go or to come across, to come up out of." kwembewale ass jive te gwa nya That boat is going across boat that goes on river this this river. (P03) I 'a kntlégu balé jive te xwégkwa (he)jumped across the road. cut jump go over on road (WS29h) save mbe melemé geg'yt... We came from the village... come in village we (WS288) kweté malé save ki g8 a nci A certain woman came from some woman came out house of them her house. (P23) f. -mbe "to go in" This is only used with dze and seems to_func- izion parallel to save plus a NPprep with mbe. pa dzembé c8 ... Then (he) went into the house...(Gr65) go in house ...mike gkc kadzembe gwu. ...her gruel is (boiling over) gruel herb goes in fire into the fire. (PSI) _ g. Two other pairs of stems occur frequently in my Corpus, although VVIlether they represent the same sort of suffixation discussed above is IPzroblematical. They are 11 and fig always used with 53 "compound" to Ineean "go" or "come home", and Seké/dzemté meaning "come from" or "go to 13I1e bush". xwé yaxwfl kwe ass kaji g8 That goat is running to go run run goat that to go home home. (PSh) ...gé’ gs gkc 1e yflmu She is coming home with the come home of her with water water. (P02) ntigu 'ya kadzemté I abruptly went to the bush. (TSl) suddenly I go bush Often this is used to indicate simply going away. For example, °° ocizemté jaxaggalé zaagkd "her husband has left on a trip." leave trip man of her 119 ma nda Eeké geg'yC.... When we returned.... (TSl) when return we 3.1.2.3 1’11, and kélé stems. Commonly used with these roots is a set of extension suffices calosely paralleling those used with §g_and,dgg. Phonologically they diiffer only in the presence of a [35/ segment preceding the suffix, for example, 13.133 , discussed above is paralleled by -akame .with a related Ineaning. Although these suffixes are most frequently used with the roots rioted above, they (1) may occur with other roots, that is, any restric- I txions are semological and not grammatical, and (2) they do not all (occur in the discourses constituting my corpus and a number have had to Toe elicited in isolation. Since they are not used with enough frequency I~ 1x: specify with any degree of certainty their exact function, they will lae simply listed and briefly illustrated here. It should further be Iaoted that both of these roots, fig and kélé occur most commonly in Inarticularized form with one of the common Vi extension suffixes de- scribed in 3.1.2.1), Thus, fuxé wu§i Bring the things. (SR2). pa kélémté wusu m8 Then she took off the lid. then took thing mouth (PSI) The suffixes elicited or occurring in my corpus are the follow- ing. a. -akawa/-akaa ~ akawa/ "take or put something down from up on something." *1 'a 'ya kékélakaa saga te gkawa I took the pot down from I took down pot on place up on the place. (Gr57) b. -akame "to take or bring up out of something." 'a kékélakame xwa kwa gweggweg (He) took the knife out of took out knife in gong-gong the tin can. (WS281) c. -akate "to bring or take up, but not too steeply" 'a 'ya kéfuakaté xa gwa I brought the millet up from I brought up millet river the river. (Gr56) d. -akayi "to take or bring down slightly" 'a 'ya kékélayi Saga laté g8 a Hunter I took the pot down I took down pot up there house of from Hunter's (Gr57) Hunter. lStarred forms here are those elicited in isolation. 120 e. -akave /-akav8 ~ akave/ "to bring or take over or across" *‘a 'yé. kéfuakave xa pelt: melemé I brought the millet over I brought millet over village from the village. (Gr56) f. *‘a 'ya kékélakambe saga mbe CC I took the pot out of the I took out pot in house house. (Gr56) Following are several examples of these suffixes with verbs other than £13 or kélé. pa kanefakaa Then (it) boiled over. (P22) then boil-down 'a 'ya kélakate I dug (it) and brought it down. (Grh2) I dig—bring §.l.3 Markers and Affixes. In Kapsiki the sememes corresponding to the dimensions of action (cf., 3.0) conceptualized by the speaker are realized grammatically by one or more of a series of grammatical indicators. These indicators take the shape of markers (free morphemes which have no meaning apart from the verb they qualify, i.e. , which are the realizations of rela- tional or circumstantial sememes), prefixes and suffixes (bound mor- phemes with fimctions nearly identical to that of markers but realizing different specific sememes) and auxiliary verbs. Unlike English, the markers and affixes in Kapsiki are often optional. That is, the verb stem or root may be used if the context is clear as to "how" or "when" the action expressed by the verb takes place or if this information is not germane to this particular expression of the action. (cf., 3.1.h for a fuller discussion of this.) In this section the markers and affixes will be simply inven- toried and illustrated. Their use in realizing the various dimensions Of action will be described. Markers, which have been mentioned in 2 .2.3.5, differ from affixes only in grammatical form. Therefore, both Will be described together and will be differentiated by the convention Of representing the affixes with a hyphen (-). The description to follow must be understood to represent the analysis of the data I have in hand. Thus, although the corpus on which it is based represents a wide range of speaking styles and includes ex- amples of the speech of several dozen individuals, it has presented some analytical problems which can only be finally checked with informants. In the few cases where the following description should be considered as 121 tentative, this will be indicated. Following are the main verb markers and affixes used in Kapsiki. (The tone markings represent the basic tone, Cf., 2.1.h.1) 5.1.3.1 '3. "Active Marker". The exact meaning of this marker is difficult to Specify since it is used in a variety of constructions. However, at least one function Seems to be that of marking verbs as active although they may be uninflect- ed (as in the imperative) or inflected with an affix giving the verb a sta- tive referent in certain constructions. La is used in the following con- structions: a. To mark the active form of a perfective verb in a simple clause.1 3 'a x81 kégate c€ They built the house. (WS350) I they built house _'_a_ g'yf késé kamené tlené mbéda We worked yesterday. (381) we came do work yesterday b. To mark the active form of both verbs in a compound clause re- lating simultaneous actions. The first verb takes a completive form and the second the imperfective. (This use of _'_a_ is relatively rare, in fact I didn't discover it for several months and then only heard it three Or four times, apart from elicited examples.) ...gezeté Y€su La nde kadzemé gwemé ...Jesus said as he was said Jesus he goes up heaven going up into heaven. (SBlO) La kégezeté da La nde katli He spoke to me as he was leaving. said to me he leaves » (Gr31) c. To mark the optative, used with an uninflected verb root or stem. La, gke kelte da lekwesa... He should take my shirt for me. he take (for) me cloth (TSl) 'a na gwenakd YOu must send (him). (Gr90) you send 3 —l.3.2 ké 'Perfective Indicator'. The label 'perfective indicator' has been used to indicate that the ké- prefix, although primarily an aspect marker in actual usage also \ lAlthough this marker occurs very frequently in my corpus it is uSed much more Sparingly in normal discourses. It is never used in con- junction with sequential indicators, nor if the verb is qualified by an aIlverb. For these reasons it occurs mainly in conversations and in short, elicited responses. 122 hats a temporal referent. That is, unless overtly marked with a time vnord, most verbs exhibiting the perfective prefix refer to actions which tuave taken place prior to the speaker's present. This prefix is used in tune following constructions: a. without any active marker ('5 a sequen- 1:ial marker or adverb), to form the participle, indicating a state Jresulting from a previous action. In many cases the reflexive pronoun gyé is used in this type of construction. newe ye tsé képesépesé You (pl.) see now! (It) is grown. see you now grown-grown (P02) kéwuba gé cs ca That house is big. (WSh32) big self house that b. Without an active indicator, to express what would be translated :into English as the 'gerund' form of the verb. In this type of usage it sometimes occurs with the potential aspect prefix te-. kanewe wundu nya kémene tlene xé They are looking at the looks person this doing work they person who is working. (Gr25) 'a kélrté gu keza newe wundu (He) turned his head in order turned head to see person to see who was coming. nya kétésé (wse67) this would come c. Without an active indicator, to express the negative perfective. ...kédepu wusu aSé 'ya wé I didn't know. (TSl) knew thing that I not kéxeve kelepé xé wé They didn't catch any fish. caught fish they not (PZl) d. With an active indicator (either 'a an auxiliary verb, or an ad- Xrerb), to express completed action in the past. (Note restrictions on tlhe use of 'a in the footnote to 3.1.3.1. above.) 'a késaté (He) stood up. (WS255) naxe kékelemté wusu mé (She) took off the lid. (P01) finish took off thing mouth rrfpam kémté gyé nde Ban: She died (CKl) died herself she e. Frequently ké- plus a verb is used to express doubt about the ifuture completion of an action or to indicate that something should follow its eventual completion. ma kédziyi genaké Yegfi... When you are there tell John».. when go-over tell John (TSl) 123 3.1.3.3 ka- 'Imperfective Indicator' Unlike 53:: this prefix rarely marks temporality. a. Unmarked a verb in the imperfective can be used refer to any taction not viewed as completed. 'The action may be in progress, not have started yet or be in the past. pa kaisre (She) will stir it. (P22) kadzegga xé te yemu They are going on the water. go they on water (PZh)‘ b. ‘52; is used to indicate habitual action. kayidé da mbeli One makes fun of me. (P25) harrass me people c. In sentences (cf., 2.3.3.2b) ka- may indicate a purposive re- lationship of the verb so marked to the first verb. kadzaa Sekwu kapa kelepe (She) is going down to market to go-down market buy fish fish. (P22) d. With te-, ka- expresses an action not yet taking place or one that would have taken place. katexi 38 a da 'ya xénde I will repair my house tomorrow. will-fix house my I tomorrow ‘ (SP1) ma nda 'a na késé nda katezeme mfi If you had come we would when you came will-eat we have eaten. (SSS) e. Prefixed to verbs marked with.:g, ka- indicates Simultaneous action in the past. (cf., 3.1.3.1b above for examples.) 13.1.3.h te- 'Potential Indicator'. This prefix seems to be a time indicator but it markes the time (of an action relative to the agent's present, rather than the speaker's. IIt is used in the following types of construction: a. By itself, that is, without other markers or affixes, to mark 1:he purposive construction in sentences with a V2 and a few Vl's in 13he PRD of the main clause. (cf., 2.3.3.2b). * dziyi 'ya vaké tenewe gada I am going over to see mine. go-over I by-here to see mine (PZl) 'a kéze gu tesé (He) forgot to come. (Gr6h) forget head to- come 12h b. Preceded by ka- to mark either future action or conditional action :in the past. (cf. 3.1.3.3d above for examples.) c. Preceded by ké— in an unmarked verb to indicate a future 'gerund'. (cf.,3.l.3.2b) 'a xé kenaké wundu nya ketemené They saw the man who was they saw person this doing going to come. (Gr25) d. Used with a number of verbs to realize a dubiative sememe.l CDhere are several forms this use of te- may exhibit, that is either Inarked or unmarked and with or without the perfective prefix. maya 'a va késé ni nda If the rain had come the house may if rain came then have fallen. naxe cs ketendla (Gr68) finish house would-fall 'e ketetsate watxé Perhaps he sneezed. (Grl?) would-cut sneeze mcslfi tenza nde He is rich. (It appears so, at rich stays he least.) (WSh89) 13.1.3.5 ya-2 'Continuous Aspect Indicator'. The prefix ya- with the object preceding the verb is used to raark.the continuation of the action expressed by the verb either in the past or the present. dzemté wume za yadé na...kageze Do you want me to talk to leave marry man wish you say you about leaving to Da ,yé marry a man. (CK1)‘ you I gwambé yaxe xé They were catching frogs. (PJl) frog catch they wupé yawupé ysmfi The water is boiling. (WS368) boil is boiling water ‘ v 1It should be noted that in several locutions the three pre- fixes, ké- _, ka- and.£g;, are used apparently interchangeably, i.e., ‘38cur in free* variation. For example, tenza, kenza and (k anza are all Ilsed to mean “perhaps. Also, the interrogative nominal where" can 'take either the Shape tema or kema. 2Possibly this morpheme is the same one used as a c0pula before Certain prepositions as in, gwu yambe yemfi "Fire is under the water" and as a relator between some emphatic pronouns and a verb, as genké ya figse ndaglya "It is this I told to him." 12S I3.l.3.6 ‘22 'Consecutive sequential Indicatork pg is used to indicate that the action of the verb marked by it eeither follows or precedes in time another action or state. It is used :in the following constructions: a. Before an uninflected verb to indicate consecutive completed eaction. (In other words, when 22.15 used, the ké- prefix is not.) dziyi xé p§_xé xeve kelepe They went across and then went they then they finished fish they fished. (PZl) pg saté pg dzembé cs {pg dexwfi... (he) got up, went in the get up went in house ate house, ate... (Gr72) b. Occasionally before a verb exhibiting an imperfective prefix. GEhis is rare since other sequential indicators are usually used in ishese constructions. IEEIkazexwfi gagké... Then she eats hers... (P22) then eats hers pé_katere Then she will stir (it). (PJ2) then stirs c. Sometimes this marker is used to indicate a state, which while Jaot following an action in time, follows the description of an action ill the narration. pa feluré mbe dseve vé Then, there is a flower in (her) then flower in hand to hand. (PJ2) pa, kweté wufé ‘32 kwagwela ta gké Then there are some trees then some tree on left of her to the left of her. ' (PS2) 55.1.3.7 me- 'Indicator of a State Interrupted by an Actionh ‘22; occurs very rarely in my corpus. This is due to the fact tkuat it qualifies a highly restricted class of verbs, namely, those IWEferring to positions of the body, such as, "standing, sitting, kneel- ing" and etc. Furthermore, it is only used when such a body state is (iescribed as being interrupted by an action. mgtleté geg'yé nda dzembé nde We were standing when he came standing we when came-in he in. (Gr35) ImgntiSé 'ya nda sé nde I was seated when he came. seated I when came he (Gr77) 126 25.1.3.8 kwa- 'Completive Sequential Indicator' kwa-, often preceded by the introducer ma, is used to indicate tune completion of one action before the commencement of another. kwafedere 'ya te x8 pa 'ya dzate... When I had wakened, I awoke I on sleep then I went-up went up to ... (TSl) ma kwasé nde naxé 'ya kédexwfi When he came, I had eaten after came he fininsh I eaten ' (SS7) (Cf.,also, 3.1.3.3a for an example.) £5.1J3.9. nda 'Incomplete Sequential Indicator'. 222 is used to indicate that an action was/is not completed (either in process or not started) when another action took/takes lglace. Egg is used in the following constructions: 7 a. Without any inflection on the verb to indicate that one action teas going on when another started. mba 'ya kawindiya nda geze nde... I continued to write when continued I write when said he he told (me)...(TSl) nda §é_'yé ni kanag-da nde When I came he was waiting. (8811) b. With verbs in the potential to indicate that an action will fleUke place when conditions have been met. ma nda tesé nde naxé 'ya kédexwfi ‘When he comes I will have when comes he finish I eaten eaten. (SS3) ma nda tesé na kazeme wusu 'ya When you come I will be when come you eating thing I eating. (SSS) (3. Before a verb with the perfective prefix to indicate that an a-C=‘bion would have happened, but didn't, if the act expressed by the IDEEJrfective had been completed. ma ya ‘a va késé nde naxe cs kendla If the rain had come if rain came then finish house fell the house would have fallen. (Gr68) <1. Before a verb in the imperfective to indicate that a not yet ffiiliished action (either not started or in process) will either precede ‘31‘ follow another action. The order is marked by the verb used to ex- Press the other action. ma nda kasé nde nda tezeme 'ya When he comes I will eat. when comes he then will-eat I (SS5) 127 3.1.3.10 keia 'Purposive Indicator’. ngins used to indicate that an action expressed by a verb is the direct result of a first action, the first action having been ex- pressly performed to bring about the second. Whether this marker is used to realize a sememe different from that realized by the purposive use of ka:_ (cf. 3.1.3.3c), is not entirely clear. It appears, however, that 5322 indicates a more emphatic or deliberate element of purpose wereas the kg; construction may serve as a simple explanation. £222 occurs in the following types of construction: a. Before an uninflected verb stem or root. (This is by far the most common use of the marker.) ...11 Sa gs keia mene tlene ta gké ...(she) goes to her house leave to house to do work of her to do her work. (PS2) teké kaxu mbeli keEa li kagiegwé... Here people grind in here grind people to go greet order to go and greet... (P22) b. Occasionally before a noun to indicate that it represents the re- cipient of the action of the preceding verb. ...kasé gké 1e keia wale a gké ...she comes with (it) for her. come she with to child of her children. (P22) gegké kaxu keEa.xu ki gké She grinds to have flour in her she grinds to flour in her house. (P22) 3.1.3.11 -akwe 'Completive Sequential Indicator'. This suffix appears to be both semantically and grammatically related to Egg; (cf., 3.1.3.8). Hewever, it is used only in pleonas- tic constructions in which the first occurence of the verb root takes this suffix, being otherwise uninflected and the second occurence be- ing an uninflected stem. Sometimes this construction occurs with a cognate object. menakwe 'ya menté tlene ta da I did my work before coming here did I did work of me a little while ago. saté 'ya meSina (GR62) come I while ago Eeyakwe nde péke pé nda teli nde He bathed before leaving. 3.1.3.12 -te 'Completive Indicator' Although this is technically an extension suffix explained in 128 3.1.2.11, it sometimes serves no other apparent function than to in- dicate that the action expressed by the verb is completed. (cf3,3.l.2.li for examples of this.) 3.1.3.13 SEXi, dla "only", 'Indicators of a Very Strong Wishh Although these two words were listed in 2.2.h.5 as conjunc- tions. However, in some constructions they function to mark the action expressed by the verb as being nearly imperative. dla 'ya meneté Sande I must do it, anyhow. (Gr7l) only I do despite Seyi dla mu kenza le wunu nya we must have the heart which only we stay with heart which Jesus gives us. (SBh) ndeke 6h Y€su gives us Jesus 3.1.h Uninflected and Unmarked Stems. As noted in 3.1.3, stems and roots are often used in Kapsiki with- out formal grammatical indicators in several types of constructions. In some cases this usage can be considered as optional, that is, where ‘the context serves to realize the dimensions of the action expressed by 'the speaker. In other cases the absence of indicators is obligatory in 'that the lack is a formal grammatical marking in itself. Several of “these usages are illustrated below. 13.1.h.1 Stative Form of the Verb. There are at least two common uses of the stative form of the ‘verb (not including uninflected stems which are found after certain. Inarkers and auxiliaries, which are not considered stative). These are (a) stems used with reference to states resulting from actions in the 1immediate speech context, and (b) roots which are normally used in the Stative form to express a general or continuing state. The following eKamples illustrate both. 61. ntigu kadzeki g8. ee. dzeki belaxc One goes into the house. leaves goes in house. yes gone in lies Yes. In the house she , ’ down lies down. (CKl) 13. mba 'ya mbe c8 kajaga... I'll stay in the house and read... stay I in house to read (TSl) ...yita nza nde butébuté ...incrder that it remain hot. in order stay it hot-hot (P22) 129 3.1.h.2 Imperative use of Stem. The imperative is formed, if second person singular, by the simple verb stem and in any other person (except first plural exclusive for which there is no form) the stem plus a pronoun. The second plural pronoun is mp which ordinarily is used as an inclusive pronoun.l zeme Eat'. (you sg.) (Grh8) sa mu Drink! (you pl.) (Grh8) kaS mu p5 Wash yourselves! (Grh8) geregé ng That he stop! (Grh8) zeme bu mu Let's eat'. (all of us) Grh9) dzegwe bu Let's go! (the two of us) Grh9) There are two special imperative stems for the verb dze "go". ggara go! (YOu sg.) ggara bu Let's go! (The two of us) mbala Go! (You sg.) mbala bu or more commonly mbigg Let's go! (The two of us) 3.1.h.3 Stem as verbal Noun. Roots and stems often function as nouns. AS such they can .fill nearly any grammatical slot filled by V1. For examples and a fuller description of this use of stems see 2.2.1.1f. 13.1.5 Auxiliagy Verbs. A number of the sememes in Kapsiki corresponding to the dimen- sions of action conceived by the speakers, are realized by the use of zauxiliary verbs. Dependent clauses built around auxiliary verbs have Iaeen discussed in 2.3.3.2c and have been noted in several examples. Auxiliary verbs behave differently. Some may be inflected 8&3 may the main verb of a clause, others, due to semological restric- tions, only occur in the stative form. Following are the most impor- tant auxiliary verbs. \ 1Whether the inclusive pronoun is to be considered first or fieczond person is, of course, a matter of convention. It can be glossed, yTou and I". It seems possible, from the use of the imperative pro- ncnun that in Kapsiki it should be considered second. 150 33.1J5.1 mba "continue", tému "repeat, last", nza "stay, exist". These verbs are all commonly, although not exclusively, found 2111 the stative form. In many constructions they serve as auxiliaries ‘tc: another verb, that is, they function to indicate the duration, the <2<3ntinuation or the repetition of the action expressed by the second 'vwarb. (cf.,3.l.h.l for illustrations of this). 3.1.5.2 in: "wish, want , like, love". fig used, with or without inflectional prefixes, indicates a (desire to have an action completed on the part of a participant in the action. kade nde kali... She wants to go... (PSh) want she go ma ke za nya tsétsé g3 'ya kafa... This man now said, "I want said man this now wish I hear (her) to understand..(P23) 'a kedé kabeté beteke (He) wanted to throw it out (Gr38) wanted throw throw 3.1.5.3 _aS_€_ "able"l. This verb is usually used with the negative marker wé to indi- <251te a nonperfective negative. The auxiliary may take either the form 51§§§_or kgfiEL, which occur in free variation, and the verb may take the imperfective prefix or occur as a stem. kasc dzegwa we (He) can't walk. (WS271) able walk not as: kazeme dafa wé (He) doesn't eat. (SSlO) able to eat mush not as: 'ya meneté we I can't do it. (WSl98) able I do not 3-1.5.1+ péxé "finish". This is very commonly used as a completive aspect marker. In I?Elcrt, the unusual tone pattern (high-high) and its frequent occurance WC>‘l.11d suggest that it has become more of a marker than verb, however, 35ti does occur occasionaly as a finite verb, meaning to finish. As an {allltiliary it is always found in the stative form. \ lEvidently as: is a shortened form of 5225' Although this type fo‘ elision is not common it is not without precedent in the language. cf-, the footnote to 3.1.3.h. 131 naxe duxwu kenefe pa nta xu kwa When the pot has boiled then finish pot boiled then put flour in she will put flour in it. (P22) (cf.,also numerous examples in text.) 3-1.5.5 dema "does not". dema is a negative verb possibly formed from the root g3 "wish" zarid a negative suffix. It usually marks negative perfective action and may be used with either the ké- or ka- prefix or as a root. kadema nde kasé He didn't want to come. (SS9) didn't he come kadema nde va 'ya pér I don't have any. (P25) , fir isn't it to I at all ‘ kédema male taga pa 'ya... If she is not there then I (TSl) not woman there than I xeeré x8 dema kese kelepe They have given up on catching any gave up they not catch fish fish. (PZh) 3 .1.5.6 naka "ripe, ready". naka is used with the negative marker wg_to indicatethat the axztion hasn't yet taken place. ama kénaka kasé wé But (he) was not yet coming. but ripe coming not (PP5) kénaka na xevé wé You haven't yet caught any. ready you caught not (P21) 3.1.5.7 Zekwa "turn around". 2ekwa is used to indicate the repetition of an action. mbeli keZekwa mbaketé biyitsa Some one has added some again. person turned added again (SB2) mbeli as: nde teiekwa kaéré kala Those people who will again people those who again turn like turn like people which mbeli 91... (8B1) ‘people which 3 -l.5.8 ié. "come". sé both as a simple root and inflected is used as a past time InnaJI‘ker. With some verbs, namely the many uses of tsa "cut", and de- SQI‘iptions of body states, §g_appears to be a genuine past marker. WjLth other verbs it appears to have a recent past meaning, in the sense ‘tlliit, "one comes from doing something." sewn ,Eé nde va gudlunu He was afraid. (WS238) disturbed him because of fear 132 késé kajene da wé She didn't answer me. (TSl) came reply me not 21 yasé nde mbisé He was sucking my sore. (WS3hO) sore come he suck 'a késé tsa 'yitla (He) coughed. (wstzo) came cut cough 3.2 SEMOLOGICAL REALIZATIONS The foregoing description of the formal grammatical properties of verbs has served to illustrate in itself how verbs may be the real- izations of action sememes. It has also shown a number of the dimen- sions of action perceived by the Kapsiki speaker and the grammatical realizations of these dimensions. In the present section I will attempt to classify some of the dimensions of action represented in the semology as suggested by the analysis of the grammar and to describe, insofar as jpossible, the Kapsiki's conception of these semological features. Four semological features related to action sememes (that is, 'which are realized as verbs or verb indicators) will be discussed to Show the relationship between this part of the semological structure and the grammatical realizations. Since, however,the actual mechanics or tactics involved in the realizations are not the issue, no reticula ‘will be depicted. In a following section (3,h) some slight attempt is :made to describe the semantic field symbolized by the features under consideration here. Thus in diagrammatically representing Section 3 the following simple illustration will indicate the general areas of con- C ern . 3 5 h WORLD OF REALITY 5'2, {; SEMOLOGY 3 ‘1’}.{ GRAMMAR Language ' PHONOLOGY SPEECH SOUNDS 133 This presentation departs significantly from the traditional (neo-bloomfieldian) approach to grammar. That is, the grammatical :features described up to this point are being classed according to the semological structure they realize rather than according to the formal grammatical properties they exhibit. This is, however, not a return to older grammars which attempted to explain grammatical phenomena by a "Latin" model through intuitive processes based on an internalized set of Indo-European language principles--although the categories they often worked within were semological. Rather, a serious attempt has # been made to specify, through the careful examination of usages in con- text, the semological units realized by the verbs and the verb indica- tors in the grammatical stratum. The rationale for this approach rests on the conviction that the comparative study of semological features of a language can have significance for the social scientist interested in explaining human behavior. Like phonology, semology, standing as it does "close" to physical reality (in comparison to the grammatical strata of language) should exhibit a greator response to and should reflect a closer rela- tionship to external stimuli than the rest of language. Thus, in much the same manner in which comparative phonology has yielded important results bearing on non-linguistic behavior, so tooshould comparative semology. A necessary first step, however, must be the isolating of comparable units. In addition to the above theoretical justification, the follow- ing considerations govern the format and the choice of content in this presentation. (1) Since most of the grammatical constructions on which this discussion is based have been amply illustrated few examples will be included here. (2) Since as noted above, some of the semology can only be finally described when a great many more data are available and some of those at hand can be checked, much of the following will be suggestive and tentative. (5) Finally, while aspects of the total verbal and non—verbal context of speech in themselves serve to realize much of the semology of events, this discussion is limited to those features realized by verbs or the verbal indicators discussed in 3.1. The four semological features considered here are: (l) tempor- ality; (2) aspect, (5) modality, and (h) roles. This classification of 13L :features does not pretend to be exhaustive and is recognized to be somewhat arbitrary and conceptual rather than empirical. However, sememes symbolizing these dimensions of language can be posited for Kapsiki and some description of them and of their grammatical realiza- tions should facilitate the comparison of Kapsiki semological structure 'with others, our own, for example. 3.2.1 Temporality. Temporality refers to the sememic representation of the time context of actions as conceived by speakers of the language. Kapsiki sememes reflect the conception of at least two different kinds of time and possibly three in the verbal expression of actions. (1) Time relative to the speaker's present. (2) Time relative to another action, and possibly (e) Time relative to a participant's present. 3.2.1.1 Time relative to the Speaker's Present. This kind of time is only optionally marked by verbal indica- tors in Kapsiki. The lack of indication, however, is of two kinds, (1) that where non-verbal grammatical features serve to realize the temporal semology, as for example total context in a gena a xéca "fable" or by a time word and (2) that in which the verb is truly neutral as regards time, as for example,often happens in descriptions. This second type of unmarked time probably reflects an extension of the principle of general and particular action which has been noted in 3.1.1.0 and which will be more fully discussed in 3.2.2. When time relative to a speaker's present is marked this is usually done by the use of the auxiliary'sg if the action took place in the speaker's past, and by the prefix ts; when the action takes place in the speaker's future. (cf.,the uses of 32; as described in 3.1.3.h) With fig other semological features such as aspect and role are realized by the verb indicators accompanying the auxiliary, thus the occurrences of késé, yasé, etc. described in the discussion of the auxiliary use of fig, (cf},3.l.5.8) With the potential indicator these features are realized'by other indicators accompanying the main verb. In independent clause coretructions, particularly in one clause discourses or in conversations, the construction composed of the verb in the participial form with an action marker can be considered to mark 135 this kind of time. However since this only occurs if the total semo- logical context is exactly right, it can also be argued that this, not the verb indicators are the grammatical realizations of the tem- poral sememes. (cf.,also the discussion in 3.2.1.3) Thus, 'a késaté "He stood up.", while realizing a past sememe when uttered as a com- plete discourse, can be analyzed as follows. The marker La_serves to realize ACTION,1’2 the prefix 3:3: PERFECT and the context SPEAKER'S PAST. 3.2.1.2 Time Relative to Other Actions. This kind of time is ordinarily realized by sequential indi- cators. Many of the sequentials realize complex sememes, that is, both temporality and some kind of aspect. Thus nda, while indicating that one action is conceived as taking place in a particular temporal relationship to another, also indicates that the action realized by the “verb it qualifies is imperfective. 532- and :2533, on the other hand, indicate that the actions of which they realize a dimension, is com- pleted before the start of a second action. ‘mg: must be considered both a sequential and an aspect indicator since it marks a state but only relative to a simultaneous action. pa and the combination 'a...ké plus 'a...ka-, may on occasion realize simple temporality but in other contexts they are distinguished from.each other in the kinds of action qualification they realize. The relative frequency of sequential indicators in many styles of discourse would appear to logically follow from.the observation that grammatical constructions tend to be verb-centric. (cf., 2.2.3.0) It would appear that some segments of the semological structure are also action centered. That is, that attention is focused on one action and that other events (action or description) are conceived as standing in a temporal relationship to it. lSememes are represented by English words in high case letters. This serves to distinguish them from morphemes or lexemes usually represented by lower case letters which are underlined when used in the text. 2These sememes are more fully discussed in 3.2.2. 136 3.2.1.3 Time Relative to Event Participants or Circumstantials. 'Whether or not this kind of time is formally marked grammati- cally and/or symbolizedl semologically, is not certain. It appears possible that the use of the participle (i.e., the EEK form) without an active indicator and perhaps on occasion with one, is used to mark the time of an action relative to an agent participant. For example, kéwuba 36 ca "The big house", 'the house sometime in it's past became big, and now is big.’ However, this can also be analyzed to reflect time relative to the speaker's present, or in many contexts, since focus would be on some particular action in the discourse, time rela- L tive to another action. At any rate, the distinction between the uninflected stative and the participle which is also a non-active form of the verb, seems to include a realization of temporality. 3.2.2 Aspect. The conception and realization of aspects of action is a highly complex, and so far, little understood problem. From the data presented in this thesis, however, at least three types of aspect can be assumed to be reflected in Kapsiki semology. They can be most easily expressed as three sets of binary oppositions of sets of con- tinua with the oppositions depicted as polar extremes. They are (1) active - inactive, (2) general - particular and (3) perfective - im- perfective. . It must be understood that the term aspect is here used with a semological and not a grammatical referent and thus differs some- what from its use in many traditional grammars. Furthermore, the exact nature of the interrelationship of these three sememes is not entirely clear. They are not mutually exclusive in a single action event configuration nor are they all obligatorily present. In fact, the notion of neutral - non-neutral may be an important kind of gen- eral - particular sememe or hypersememe. In depicting the semological network graphically it would 1The term symbol is used to express the relationship between the semology and reality. This is to distinguish this type of rela- tionship from that enjoyed between strata which is referred to as one of realization. 137 jprobably be necessary to posit at least two strata, that is, a hyper- sememic and a sememic to represent these relationships. (cf., Lamb 1966, pp. 31-33). This is to say, it appears that some or thnu un- pect sememes are arranged in some type of hierarchy. However, since our discussion is not primarily concerned with the mechanics of the system it will be limited to simply pointing out how parts of this network are realized grammatically. 3.2.2.1 Active - Inactive. In the grammatical sketch three types of inactive verb forms A?- are noted, verbal nouns (2.2.l.ld), stative verbs (2.2.2.0 and 3.h.3.7) and participles (2.2.2.0). Since all of these can be the realizations of events (although in some cases, particularly with verbal nouns, they can also be the realizations of referential sememes functioning as participants or circumstantials) and since all can be shown to ex- hibit morphological relationships to active verbs it seems fruitful to consider their occurences in some contexts as the realizations of as- pects of action as conceived by the speaker. It might be that the difference between the verbal noun and the stative or participle is one of degree of imputed activeness. When the stative is used this seems to indicate that attention is being focused on the continuation of a state which is existing, either re- lative to the speaker's present or relative to a point in time rela- tive to an event in the discourse. Whereas, the use of the participle focuses attention on either the quality a state which has resulted from an action confers upon a participant, or to the fact that an action has resulted in a state. Thus, at least part of the distinction between the following two uses of the non-active verb, is explained. 'a xé kenaké wundu nya késé They saw the man who was coming. they saw person this coming (Gr25) mentise 'ya nda sé nde I was seated when he came. seated I when come he (Gr77) These examples illustrate verbs which may occur with or with- out active indicators, although the few which can occur with me- realize a special sub-class of sememes. Class IV verbs which always occur in the stative form (cf., 2.2.2.h) present a different problem 138 of semological analysis. They can be considered the realizations of events other than actions or they may be analyzed as the realizations of inactive aspects of action events with no corresponding active sememe retained in the language. (The status of the grammatical units, i.e., their classification as verbs, is at least in part predicated upon the fact that they occur in constructions with the nominative form of the personal pronoun.) The realization of the sememe ACTIVE as a dimension of an action has been described and illustrated at several points in the grammatical analysis and needs no further comment here. 3.2.2.2 Perfective - Imperfective. At some points the conception of perfective and imperfective aspects converge with the aspects just discussed. Participles, for example, may be viewed as the realizations of perfected actions re- sulting in a state, or as the state resulting from a perfected action. However, stative verbs appear to realize states with little regard as to whether the action resulting in the state (if in fact one is viewed‘ as causing it) is perfected or not. There are several means for realizing perfective action with verb indicators apart from their contextual realizations. The most commonly used is ké;, either as a participal formant or with an active marker. The others, include the uses of kwa-, -akwa and -te. (cf., 3.1.3) Perfective action is conceived as being finished, either in the past or at any other point in time. Imperfective action, on the other hand, is viewed as not being finished either from the perspective of the speaker's present, or from another vantage point, as for example, is the case when a second event is realized as a purposive verb in a kg! construction. (cf., 3.1.3.3c). Imperfective aspect indicators may be of several different types and the distinctions serve to realize significant relationships in the semological hierarchy. These indicators include kae, te-, ya-, nda and their several combinations. (cf., 3.1.3) nda and ka- seem to be the realizations of generalized imperfective actions, the former used in sequential constructions. The realizates of the potential indica- tor te- and the continuous indicator ya- have been discussed in 3.1.3.h 139 and 3.1.3.5 respectively. The auxiliary verbs (cf., 3.1.5.1) which function grammatically to qualify verbs, should not be viewed as realizing the imperfective aspect of action since they are normally used with other imperfective (or, as the case may be, perfective) indicators. Rather they serve to realize the intensity, the duration and etc. of the imperfected action. (This, of course, can be also considered a realization of aspect.) 3.2.2.3 General - Particular. The notion of general and particular as it applies to the con— ception of actions, has been alluded to a number of times (cf., 3.1.1.0, 3.1.2.1, 3.2.1.1). As with the question of neutral and non-neutral in respect to temporality, as well as the question of imperfective dis- cussed above, this could best be depicted graphically by the positing of hypersememic and sememic strata in the semology. No general sememe of particular aspect appears to be realized and thus if posited must be done so at a hypersememic level. (Possibly the grammatical unit 'verb suffix' could be considered the realization of such a sememe, but this seemsrxtto be a fruitful approach, since each suffix does realize a certain kind of particular.) The most pervasive kind of particularization seems to have to do with spatial context of action. In class I verbs this is expressed relative to a participant (usually the agent). In class II verbs the spatial context relative to the agent is included in the semantic field symbolized by the sememe realized as a root, and the particular- ized aspect refers to the terrain over which the action unfolds. The several exceptions to this do not necessarily vitiate the argument but might, with additional data, suggest a refinement in the analysis. Thus :tg_shou1d perhaps not be considered an extension suffix and thus a stem formative but a verb indicator, as well as, possibly, :kg and :akg and 2252f (cf., 3.1.2.1) Virtually all of the other extensions realize, at least in some contexts, spatial sememes. Thus, :kg refers to action progressing to- ward the speaker, :52 action away from the speaker, :yi_indicates action downward in some semological contexts and over in others, :223 refers to seperation from the agent, etc. This is also true of the suffixes lhO found primarily with.kglé and f§,that is, they realize spatial dimen- sions of the act of taking. As has been indicated in section 3.1, the sememe GENERAL is usually realized by the unextended use of the verb root. The various particularized sememes are realized by the stem formatives described in 3.1.2. 3.2.3 Modalit . Ch of the things distinguishing human language from animal l speech systems is the ability to express moods, or as Greenberg has termed it, the quality of multimodality. (Greenberg 1968. 9) The exact parameters of modality are not easy to define and certainly it is empirically difficult to distinguish from the expression of other dimensions of action. As a semological phenomenon, modality, here, refers to the subjective commitment to the event on the part of the speaker. Precisely how many different modes are conceived by the Kapsiki as distinguishable from each other and from other dimensions of action is impossible to determine from the data at hand. Probably the degree of subjective commitment to an action can best be viewed as a point on one of a series of continua. That is, a speaker may at one extreme be opposed to an action taking place and at the other extreme desire it so strongly as to command it's accomplishment. In between can be ex- pressed various kinds of interest in seeing the action completed in- cluding disinterest, suggesting it be done and hoping or wishing for it. This, of course, represents only one of the continua into which the sub- jective relationship to an action can be mapped. Another is that of dis- belief - positive affirmation. Various points upon each of these continua, it appears, are realized by Kapsiki verb indicators. These are briefly discussed below. 3.2.3.1 Imperative - Interdiction Axis. The realizations of five seperate sememes have been identified which can be viewed as occupying points in an imperative - interdiction continuum. They are, (a) imperative, (b) emphatic desire, (c) subjunc- tive, (d) optative and (e) negative imperative. Each of these is realized by a different set of verbal indicators in the grammar. Furthermore, it lhl should be understood, that these labels have been more or Jenn arbi- trarily chosen to symbolize posited semologicalstructures distinguishvd formally in the grammar. a. Imperative. The imperative expresses the strongest desire on the part of a speaker that an action be performed.1 As indicated and illustrated in 3.1.h.2, the imperative form is used to command a person directly (realized as a verbal construction composed of a stem and a first or second person personal pronoun subject) or to express a very strong . I wish that a third party perform the action (realized by a verb stem “ and a third person subject.) There seems to be no conceptual differ- I ence in the semology between what would be glossed as a hortative or a subjunctive in English, and what is realized as imperative. At least there is no formal grammatical distinction, the root or stem of the verb with an appropriate subject, which in the case of a second person singular pronoun may be optionally ommitted, being used as the realiza- tion. (cf., 3.1.h.2 for examples of the imperative.) b. ,thative. The optative is used with reference to the grammatical construc- tions built around a verb marked by the auxiliary verb gg. This con- struction is used to realize a wish on the part of|a participant in an event. The participant may or may not be the speaker as well. As illustrated above, if other dimensions of the action are realized gram- matically, this will be done by the use of indicators affixed to or marking the auxiliary verb. (cf., 3.1.5.2 for illustrations of the optative.) c. Subjunctive. The subjunctive mood is realized by the uninflected verb stem marked with the active marker 'a. Semologically the subjunctive sym— bolizes a suggestion on the part of the speaker that an action be per- formed. The suggestion may be made directly to the person involved 1Perhaps a still stronger form can be posited. The imperative may be used with 5; which appears to act as a strengthener or emphasis marker. This kind of construction could be interpreted as realizing a sememe standing at the end of they hypothetical axis described above. However, it has been treated as not being distinct from the imperative. 142 or it may be made concerning a third party. The distinction being marked by the subject of the verb in the grammatical construction. (cf., 3.1.3.1c for illustrations of the subjunctive.) d. Emphatic Wish . The emphatic is used to realize a feeling of 'oughtness' or obligation on the part of a speaker or participant as conceived by a speaker. It is indicated grammatically by one or both of the markers dla and.§§yi, (cf., 3.1.3.13 for examples of the emphatic use of dig.and‘sgyi.) It should be understoon that although these three moods, op- tative, subjunctive and emphatic, are distinguished on the grammatical stratum, they may on occasion all be realizations of the same sememe. Furthermore, there may be features other than modality-~as conceived here-~which are realized by these grammatical forms and which may thus make the choice of one or the other obligatory in some contexts. This analyst's competence in the language is not great enough to make such specification at this point. e. Negative Imperative. The negative imperative (discussed in 3.1.3.1h--notice par- ticularly the uncertain assignment of meaning to mpg) appears to realize a strong wish or command on the part of the speaker that an action not be performed. It is realized by the marker mpg plus-the participle form of the verb. It should also be understood that negative can be used to ex- press a wish or hope that the action not be accomplished. For example: asé kwe asé kade kanza...wé That goat doesn't want to stay. able goat that wish to-stay not ‘ (PSH) The auxiliary verb dema (cf., 3.1.5.5) seems to be used to realize this mood as well, although in some semological contexts it serves as a negative indicative marker. 3.2.3.2 Disbelief - Positive Affirmation Axis. The semantic relationship between negatives and interrogatives in African languages has been noted by other linguists. (William E. Welmers presented a paper on the subject as it related to Igbo in Ann Arbor in 1965.) In Kapsiki there are contexts in which an interroga- tive marker-~either an intonation pattern or a marker—-is used 1&3 apparently to signal negation. Furthermore, it appears that from the semological stratum, grammatically marked negation and interrogation can often be viewed as the realizations of points on an axis indicating degrees of subjective interest in the action on the part of a speaker. This subjective interpretation can range from conviction that the action did not, is not or will not take place to the positive affirmation that it has, is or will. In between can be expressed various degrees of dub- iousness. Thus on this continuum action can be conceived as not happen- ing, perhaps happening, hopefully happening, not known as happening, and affirmed. Some of these moods are realized grammatically by the verbal constructions (a) negative, (b) dubiative, (c) interrogative and (d) affirmative. Other degrees of doubt or affirmation can be realized con- textually or by non-verbal constructions. a. Negation. As has been discussed in sections 3.1 and 3.2.3.1, there are a number of different grammatical forms negative constructions may take. For the most part they can all be subsumed under the heading treated here, that is, the realization of semological modality. The variations reflect realizations of other dimensions of action. Thus, while wé realizes general negation, a§§_plus fig usually realizes non-perfective negated action, £322 negated perfective action and mpg imperative nega- tion. (cf., 3.1.3 for examples of all these constructions). In addition to these negative constructions the interrogative marker nag is frequently used to realize a type of negation in that it marks a question expecting a negative response. For example: pfikwfi teli gké depu wundu nee‘ Maybe it (the road) goes a far will-go it knows person no? long ways. (P03) b. Dubiative. The grammatical construction labeled dubiative is formed by the use of the verb Egg "stay, rest, remain, etc." with any of the verbal prefixes (ka-, ké-, te-) discussed in 3.1.3. (cf., especially footnote to 3.1.3.hd). This is, of course, only one of several possible realiza- tions of the sememe DUBIATIVE. Others include the use of the verb ggpg "know" plus the interrogative marker Egg as illustrated immediately above, and apparently the use of the potential indicator or contextually, as perfective. Thus: lhh 'a kétetsaté yitla nde Perhaps he coughed. (Grl8) did-will—cut cough he c. Interrogative. The several interrogative markers described in the grammar, y}, nag, w§§_and possibly wfi (identified as a clause terminator expletive) may be used to realize a question on the part of the speaker as to the actual performance of the action symbolized by the event, or a question concerning a participant in the event. This, plus the fact that various degrees of doubt may be expressed, explains the different realizations of INTERROGATION. Thus, for example, yi usually marks a question re- garding the actual performance of an action or existence of a state. mba te gweme yi. Is she living? (WS2h8) rests on life? 'a na kéxite ggameggame yi_ Did you fix the net? (WShl3) you did-fix net ? wee ordinarily indicates that the speaker is entertaining a question with reference to a participant or circumstantial in the event configuration. wa yamené na waa What are you doing? (WSth) what is-doing you ? wa nde sé waa Who came? (W8226) what one come ? né§_and HE: the latter in various combinations with other ex- pletives described in 2.2.5.3c, are used to express mild degrees of questioning, if not, in fact, rhetorical questions. If a response is to be elicited it will be negative with Egg and affirmative with E2! The following excerpt from a conversation between two individuals (Kwadade and Luc) illustrates the use of E3: (K) pa mbeli xkaké wfi Then one called her, O.K.? then person called ? (L) e Yes . (CKl) d. Affirmation. Positive affirmation of the performance of an action is realized by the lack of any indication to the contrary. This has been illustrat- ed throughout the thesis. The affirmation may be emphasized in a number of ways, but since these techniques, with the possible exception of the use of hi. (cf., 2.2.5.3d), are not realized as verb indicators, they 11+ 5 are not discussed at this point. 3.2.h Roles. It is not possible at this time to provide a complete dCSvriy- tion of the semological structure relative to roles in Kapsiki. How- ever, selected features are discussed which reflect significant as- pects of the relationships conceived to exist between participants and events in the semology. Roles, for the purposes of this discussion are therefore, defined as the sememes symbolizing the relations between actions and actors or objects involved in the action, excluding the FL- speaker whose relationship is symbolized as modality. These relational .-.L'. liz'gv'a‘h .. sememes are realized grammatically by a number of phenomena, including, subject - verb, object - verb, indirect object - verb constructions, ("" transitivity, passiveness, reflexiveness and cognate object construc- tions. A thorough and accurate understanding of these relational sememes is a necessary precondition to understanding the Kapsikis' view of the universe or universes in which they live. However, such an understanding can probably only come with complete enculturation and this includes the acquisition of a high degree of competence in the language. Lacking this, the analyst can hope to make some general hypo- theses which should at least provide accurate hints at the nature of these phenomena, i.e., the nature of the relational sememes. Several of these broad hypotheses, stemming from the analysis of the grammati- cal structure presented up to now, are adduced at this point. Although the conclusions presented here must be considered as somewhat unrefined hypotheses, since they are not predicated primarily on intuitive pro- cesses, they should be amenable to verification or, at least, refine- ment by the analysis of new data. In choosing a format for discussing roles, as defined above, it is impossible to delineate categories which admit to empirical dem- onstration of mutual exclusiveness. However, the two closely related features, (1) transitivity and (2) the characteristics of participant - event relationships, while only conceptually distinguishable, merit a brief discussion. 146 5.2.4.1 Transitivity. To attempt a universal classification of verbs (grammatical units) on the basis of transitivity - intransitivity appears fruit- less. Most verbs seen to be neutral in this respect. That is, de- pending on the particular context of the semological realizate, they may be either. In the semological stratum, on the other hand, events must be conceived as having recipients or beneficiaries, or not. The lack of such a participant may be realized by: (a) an intransitive verbal construction, (b) a noun with a general realizate as object, (o) a construction with a cognate object, (d) a verb with a general realizate in which case the event is realized as a noun filling the object slot, or (e) some kind of reflexive object construction. Option (a), while theoretically possible with all verbs, and actually occurring with a number, is rare. This would seem to suggest that the normal tactic pattern dictated by the grammatical strata is characterized by a V0 type construction. Options (b) through (e) rep- resent techniques used by the grammar to achieve conformity to this pattern in the realization of event configurations with no recipient or beneficiary participant. The following examples should illustrate these usages. (a) Intransitive Constructions. 'ya kantiffi I spit. (Gr?) ma kasé nde kaseme 'ya He comes while I am eating. (SSM) comes he eat I (b) General Noun Object. ma nda sesé na kazeme wusu 'ya Whenever you come I am eating. when come-come you eat thing I (886) Note: this last example realizes the same sememe with the V0 as does the V in the second example under (a) above. However, it should be understood that the noun 335g is in many other contexts used to realize a different type of sememe. (c) General Verb Predicates with Object Realizing the Event. katsa 'yitla nde She coughs. (Grl7) cuts cough she 'ya tsa bamba I swim. (On top of the water.) I cut swim (GIB) 11.7 (d) Cognate Object Constructions. 'a képeké pg. He washed. (wssh7) did-wash wash venexwfi yavcnexwfi nde He is vomiting. (Grfi) vomit is-vomiting he (e) Reflexive Object Constructions. Reflexive objects may be of two types. They may be nouns realizing the sememe BODY or part of the body of the agent participant or they may be the pronoun pg, 'é zewé kentle n§_ The rope broke. (WS376) - “J rope did-break itself 'a gena kénéxéve geva... The news spread...(WS399) word divided body The above examples illustrate the realizations of event con- figurations without recipient or benefactive participants. This con- clusion follows from the analysis of the contexts in which they were used and comparison to other examples. The same type of construc- tions can also be used to realize such participants. This is even true of the intransitive constructions illustrated under (a), since certain appear to realize in addition to aspect the presence of a recipient participant as well, for example: 'a kezezexwfi He ate everything (of several things?) ate-ate all (Gr27) It would seem from the above illustrations that the presence or absence of an object may reflect either the tactics of the grammar or the semo-tactics. Since no classification of verbs at the gram- matical stratum suggests itself for explaining this, the solution may lie in classes to which the sememes are assigned by the language. Thus, there are evidently certain sememes which can be positied as ex- hibiting allo-semes in their grammatical realizations. The sememe EAT for example, may be realized as: ‘3322 (that is, a verb root) or £223 ;!g§§ (verb root plus object). On the other hand, it is entirely possible that there are cer- tain verbs, pg "wash" for example, which are required by the tactics of the grammar to take an object and when no recipient participants are symbolized in the semology, a cognate construction results. At any rate, from an analytical point of view these are trivial problems which 1&8 should yield solution with the application of additional data. The class II verbs deserve special mention here. The roots used to realize actions of coming and going, realize both a goal participant (distinguished from recipient by the kind of relationship it enjoys to the event) and the event itself. Therefore, they are not normally found in constructions with beneficiary or recipient participant realizations. They thus fall outside the general consid- eration of transitivity - intransitivity discussed above. 3.2.h.2 Characteristics of Participant - Event Relationships. The above discussion involved the question of the presence or absence of recipient and benefactive participants in the semology. The characteristics of the participant - event relationships and their realizations is the wider tepic alluded to here. Participants usually symbolize entities in the world of reality. These may be animate or inanimate objects or abstractions. Reflecting the nature of their relationship to events, participants fall into two general classes. Those which are involved in the performance of the action (agent and causer participants) and those which are in some sense the target of the action (recipient, benefactive and goal participants). Partici- pants are distinguishedirom circumstantials by being more directly in- volved in the action and thus connected to the event by one of the valences referred to above, such as, agent, benefactive causer, re- cipient or goal. Participants are normally realized as nominals or NP's and when pronouns, the nominative case is used with performer participants and the oblique case with target participants (cf., 2.2. 1.2). Exceptions occur in the speech of certain dialects where pkg 'third, singular, oblique' is used optionally in placeci’ nde ~ ¢ 'third, singular, nominative' (cf., example in 3.2.3.2a). The zero realizations observed for third singular and second singular in the imperative are not considered exceptions but allo-morphs used in free variation. Normally, agent participants are realized as nouns filling a subject slot in the clause, beneficiary participants as indirect ob- ject slot fillers, recipients as direct objects and goals by the verb root with class II verbs and possibly as extension suffixes with 1M9 some class I verbs. Still another construction which can be consider- ed as the realization of role configurations in the semology is the purposive construction discussed before (cf., 2.3.3.2b). For the most part purposive participants are themselves event configurations. How- ever, as has been noted, they too may be referentials realized as nominals. The distinction between causer and agent participants is dif- ficult to demonstrate empirically and the following analysis must be considered highly tentative. Causer roles refer to the participant - a 71 event relationship realized by what would be glossed as a passive construction in English. These constructions, marked by the use of the verb laka "by", are very rare in my corpus. For example: laka gegké ndeke da geva xwa The knife was given to me by by he! gave me body knife him. (Gr79) (Literal translation: The knife gave itself to me by him.) Here the object of the verb is the realization of a general referential sememe (geva) standing in recipient relationship to the event while the subject of the verb is the realization of a partici- pant standing in an agent relationship to the verb. Another partici- pant stands in the causer relationship to the event. This latter is realized as a nominal in the emphatic form, marked as causer by func- tioning as subject of 12kg standing in a logically dependent relation- ship to the main clause.l (cf., 2.3.3.3 for a discussion of logically linked clauses.) Thus the semological relationship can be graphically repre- HIM-—-—-—(causer)-————————-GIVE (ben)————————-ME sented as follows: r tar—m 1There is a possible alternate analysis of this type construc- tion. laka could be considered a preposition and the phrase formed by it a NPprep realizing a circumstantial to the event. In this case it would not be necessary to posit a causer sememe. However, since laka doesn't realize instrumentality, since it does take the nominative form of the pronoun and since HE appears to stand in a very intimate rela- tionship to the event, the analysis presented in the text appears pre- ferable. 150 Roughly speaking the realization rules would include: 1. causer I} laka plus subject in the emphatic form (pro. third, sing.) 2. benefactive R indirect object filled by pronoun, first, sing., oblique. 3. agent R subject filled by noun. 1+. recipient R'object filled by geva. 5. participle liby ké—. 3.3 A SEMOLOGICAL NETWORK 3.3.1 The Network I present new a preliminary attempt, on the basis of the fore— going analysis, to represent graphically the semological network under- lying a portion of the narrative discourse, TSl, alluded to several times in the thesis. This narrative was recited by Mr. Luc Sunu who worked for me as a general assistant. One afternoon he came up to my office and, after he had sat for a while trying to work my typewriter, I asked him to tell me what he had done that day. The following dis- course is the result. (Cf., 3.3.3 for the meaning of the notations used in this net- work.) 151 (7) Eventl Configuration Event Line (Speaker) (1) L?C————— ag——+—AWOI(E- — ----1oc— - — — - — -—EARLY MORNING . @\ I (2) 1'- ag —— GO-UP\— -— —loc -— - -HOUSE— - —- -gen— —— — -LEWA I I @ I:"~prp---GREET- - - -rec ----- o I (5) ?* ag COME-DOWN I . (II) I MARIE ag——— GRIND I —-u-—- I .— I , I“ @ (5) II.- 4. age SAY———___ ree____.[ QUOTA‘TION] . ..__../ = ' I (6) TL 1 3&7 SAY—————rec 4 I I o———— bn @ I l I I I as (8) : L-_—~—_ag‘——:———- GRIND rec , ’ rFLOUR I : “~““-gen2”’ (9) ¥ ' ag—————————TAKE=::-—————rec HOE ' I ‘ \ ‘ ‘~man ------- SUDDENLY (10) ‘IL is ag Gog: - ....1oc_ .. - - _ --BUSH I I ‘2: \ \ ‘ ~ man ——————— SUDDENLY (ll) %1 I ag HOE rec-—FURROW—v~--num-‘----ONE : I 1The numbers of the Event Configurations correspond to the numbers of the clauses in the discourse (cf., 3.3.2). 152 Event Line LUC MARIE m a T1 W“ IIIII A, m1 m m . nu mw u H S TIIIIIJ a """" III'II|'II1|"I u w _ a. a . Y . . . m . L... . 1E . M _ n. _ vv IIIIIIIIIIIIII _ m I T as . T . _ . m . . . It I _ . . c x M c a a a a. e a a . m m ,1 r 11 r r. r x c f. .1 . x _ e g . .\ . .r a _ _\ . we \ _ _ “w c @ F @ .® 9 ® ® m 9Te T N T a .3 a L T p a a ax I m I a . no r1 mm T1 TFI T. A” nu Tu . m H S D M T T m E mu. m @\ Ami/Q ... a a Q Q ® C C L m g gs no 2 me me me we we we m we we a a a a 1-1-137: :1 ...... _'lIIII.T Ilifi rllLvll lllLvlkrllLvlllLr lllll LvllLleL.IIII m m m w m n m m, m. a m m m.- 153 LUC MARIE Event Line (Cf., ‘ “ 'P, 21) I l I | I I I Q 1‘, I (25) FINISH—_— rec ' I I I t q n I I.- I bn I (27) e : ag SAY— rec ———{ QUOTATION] I .2... . I—bn;Q-— (28) ,L i ag——-——-RETURN ' l - . I (29) ' :.__. ag ——-——ANGER ' - ® I I - , . ® a L _— __ (so) I a ag KNOW rec THING I l : | emph ® I I (51) I f 8g KNOW T— reC .4, a \ , I l ‘ gen --------- A : conj ' Q l I I __ ... _ _ .. _ ... ._ _ _ _ _- I (32 ) I r— ag STAY 10c HO'L SE I (33) it I ag ——GET-UP-— - - - man ——————— SUDDENLY ' \ I \ l I .prp- -;(’iQ-QP— --loc— --HOI\J\SE ' | ’prp/ éep I , ’ ‘ \ _ _ +—ag 1— TELL\ rec— WORD---rel-«-[ . . . . ] HUN:TER ' ‘ I\ben 4 I I I I I t I I I I | l 15h LUC MARIE Event Line HUNTER T I l I I I l I‘i:i{\ /g:::> I ‘5“) It I “1 *GIRKT 5 \ : I ‘\ rec eMONEY l \ : l ‘prp---BUY-———rec———————-MILLET . . an I l (55 ) a I ag COME-DO --.1oc---OVER-HERE- --dir— - “DAVID l I I I I Q I I I I (36) e 4, ag—-———TAKE —————rec TIIING- --pos- - - II I I I . I (37 ) L . ag———WRITE 155 5.3 A SEMOLOGICAL NETWORK. 5.5.2 The Discourse (TSI) (Recorded August 10, I967) lmesina maské-maské kwafedegé 'ya te x52 pa 'ya dzate recently morning-morning after awoke I on sleeping then I go-up g8 a Lewa kanjewe Lewa. 5ma kwasiyi 'ya nilL kaxu Mari. house of L. to-greet Lewa after came-down I thus, grinding Marie 5ma pa 'ya §e ni xwégkwa tedzemté mu nda Xu na tsétsé- said I to her thus, when will-go-to-bush we grinding you now nya was 6 a pa 'ya Es 7késé kaiene da wé 8mba nde ? thus said I to her (past) answer me not stay she kaxu xu ta gké 9nzapé ’ya te wude a da lOntigé 'ya grinding flour of her thump! I on how of me leftl I kadzemté ll ma naxe 'ya ketereke temé kwetéu l2pékwé go-to-bush .(intr.) finish I hoe-in-furrow furrow one appeared! nde le dafa 15pa 'ya gave kweté wuza lLLpa 'ya tereketé she with mush then I cut-out another cultivate then I hoed-into-furrow l5nti§é 'ya gena dafa 6pa 'ya dexwfi l7pa nde guke g'yé seated I next to mush then I ate then she dipped for us yemfi 18pa geg'yé séxwu l9mba geg'yé wuza 20mbé geg'ué wuza water then we drank continued we cultivate cont. we cultivate 21tené wuza asé bexwé 22pa 'ya belexe da mbe xa 23pa 'ya remained cult. that little then I laid down me in millet then I 25 Eeneké mpi 2hrpa 'ya sate pa 'ya kedagte wuza a§e. ma ké rested-a-bit rest then I got-up then I finished cultivate that said Ea 'ya are g'yé kéwuza 'iya kadé kadziyi Sire 27ke make to I when we cultivated I want to-go-over Sir said if dziyi na 'a na genaké Yegu kadzemé Sedakwuté 'a go-over you should you tell Yegu to go—up to Sedakote should kelte da kekwesa ta da va Zagé Zera wa pa 'ya s6 take for me shirt of me to John Zera Thus said I to her 28 .— 2 ma nda seke geg'yé 9naxe kexwepeté wuné éokédepu wusu a§é while returned we finish tied heart knowing thing that ‘ya we Slgegké nde depu wusu a gké are 32mba nde ki g8 I not it is she she knows thing of her when stay she in house 156 33ntigu 'ya kasaté Kadzaté gs a Hunter Shpa Hunter ndeke abruptly! I get-up to-go-up house of Hunter then Hunter gave da céde kapa xa be'a keta gena nya genaks as Timutiy6 me money to-buy millet anu for word which said to me Timothy kageze nda gena ta kanferans kadza Berexé kamené nda to-tell to him news of conference to-go Bourha to do then siyi da 'ya lakake ea Davité 36pa 'ya kéléte wusu ve came down me I here to David then I took thing at his 57mbé 'ya kawindiya continued I writing. Translation: Early this morning after I woke up I went up to Lewa's house to greet him. When I came back home Marie was grinding. I said to her, "Why are you grinding now since we are going to the bush? She didn't answer me, she just kept on grinding her flour. I grabbed my hoe and left for the bush. I had just finished hoeing up one furrow when she appeared with some mush. I marked out another furrow then I hoed it up. I sat down next to the mush and I ate. Then she dipped some water and brought it to us and we drank it. We continued to cul- tivate for some time. When there just remained a little bit to do I laid down in the millet and took a little nap and then I got up and finished it. She said to me, "when we have cultivated, I want to go to Siir." I said, "If you go you should tell Yengu to go up to Sedakote for me and take my shirt to John Sera." When we came home she was angry. I don't know why. She is the only one who knows. She stayed in the house and I got up to go to Hunter's house. Hunter had given me some money to buy corn and (I went up) to talk to him about the conference in Bourha that Timothy had told me to discuss with him. Then I came down here to David's and took this thing (typewriter) of his. I wrote for some time. 157 5.5.5 Glossary of notations. ag agentive pos possessive bn benefactive prf perfective cmp completive prp purposive dir directional pst past dur durative qtn quotative emp emphatic rec recipient gen genative ref referred to imp imperfective rel relative man manner sta stative neg negative sue successive nom nominalise tmp temporal num numerative qua qualificative Solid horizontal lines (‘— ) relate participants to events. Broken horizontal lines ( ---------- ) relate circumstantials to events or to other sememes. Solid vertical lines relate events in the event line. Broken verticallines indicate the continuation of a partici- pant through the discourse. Horizontal dashes C43) on the event line indicate a change in subject or arena. Heavy dots (°) indicate that the participant introduced earlier is involved in that event configuration. Brackets ([1) indicate that material has been omitted, either a quotation or a relative construction. These materials could be de- picted as existing along separate Event Lines standing in the rela- tionship to the main event indicated by the connecting line. Event and Referential sememes are depicted with high ease English letters. Events not situated along the main event line are underlined. Relational and Dimensional sememes are represented as small case English abreviations, the former imbedded in lines and the latter encircled. 158 5.5.h Realization Rules. The reason for including this network was to depict graphi- cally the semology of actions as structured by the Kapsiki language. This is done to illustrate the points made in the preceding disvus- sions. It should be kept in mind that a two dimensional representa- tion is not adequate to reflect the multidimensional network concep- tualized by the speaker. However, most of the salient features of action symbolization as events can be mapped into such a diagram. The rules listed below indicate how this network is realized in the grammar. Technically these are semo-lexic and semo-morphic realization rules. However, since the discussion to this point has predicated a three strata model, with few exceptions, they may be considered semo-grammatical rules. One of the exceptions is the event sememe ANGER, (#29 above). S/ANGER P» L/tie-one's-heart If. M/xweta wune. In this case both semo-lexic and lexo-morphic realiza- tion rules are necessary. No rules will be necessary for the realization of particu- larized action. This type of aspect has been included in the network as hyphenated to the representation of the event, where a meaning has been identifiable. Thus in event #5 COME represents the event and -DOWN the particularized aspect. This convention has been adopted to simplify the presentation and eliminate the necessity for depicting a hyper-sememic stratum. Following are the most important realization rules: Referential Sememes are Realized as: M/noun: 1) when first occurence other than SPEAKER-ag. 2) when releated with S/ref. M/pronoun nominative: 1) bn of SAY-qtn. (cf., 2.2.1.2 for gramm. rules) 2) ag if non-first occurence. 5) SPEAKER-ag first occurence. M/pronoun oblique: bn, gen, rec non-first occurence. (cf., 2.2.1.2 for grammatical rules) 159 Event semons are realized as: M/verb roots M/ verb root ..¢ 1 when found related to a referential semon of manner. (cf., 2.2.5.1a) Relational semons are realized as follows: a8 bn rec gen man tmp num dir pos PTP ref rel #3 43 J” #3 é” £3 £3 £3 £3 £3 é” £3 2” qua Clause SUB slot (cf.,2.5.2b) Clause IOB slot (cf.,2.5.2d) Clause OBJ slot (cf., 2.5.2e) NPgen (cf., 2.5.l.la) adv slot in VPadv (cf., 2.3.1.2) Clause TMP slot (cf., 2.5.2e) qual slot in NPqual (cf., 2.3.1.1b) verbs (cf., 2.2.2.u) verbu (cf., 2.2.2.h) kaV ~ keza plus Cldep (cf., 2.5.5.2b) M/asé Relative Dependent Clause (cf., 2.3.3.25) qual slot in NPqual (cf., 2.5.1.1b) Dimensional semons are realized as follows: Simple emp dur imp neg prf £3 £3 é” $3 é” 4U qtn sta 1% Complex sue prf sue cmp suc imp prf pst emphatic pronoun (cf., 2.2.1.2b) M/mbé (cf., 3.1.5.1) kaV (cf., 3.1.3.3) M/wé (cf., 2.2.3.5 and 5.2.5.2a) keV (cf., 3.1.3.2 and 3.2.2.2) M/ke with agent R 5rd. pers. sing. M/Pa elsewhere (cf., 2.2.2.5) verb stem 3' M/pa...Vstem2 (cf., 2.2.5.5 and 5.1.5.6) I: kan (cf., 5.1.3.8) f; nda v (cf., 3.1.3.9) :3 M/késé...v (cf., 5.1.5.8) lAllomorphs when indicated occur in free distribution. 2With S/sta no action marker ('a or adverb) is realized, othrr- ‘Wise one is. h. flu." 160 5.4 SEMOLOGICAL REALIZATES . This section represents simply a summary and some very pre- liminary hypothesesconcerning the relationship between the semologi- cal structure and the world of reality. In the introduction to section 2 it was suggested that language can be viewed as a coding device, standing between two aspects of physical reality. I have gone to some pains to describe the nature of the Kapsiki version of that device. It consists essentially of a set of principles for creating vocal symbols relating to segments of the world of exper- ience. Part of the thesis of this dissertation was that being able to specify these symbols would shed light on the segmentation and classification of this world. In other words, describing the lan- guage would indicate how a speaker of the language conceives the universe he lives in as structured. It was recognized in the attempt that any results obtained would be tentative and highly hypothetical. This is due in part to the level of competency the analyst has attained in the language, and in part, to the impossibility of verifying conclusions. However, the effort has been deemed worthwhile for several reasons. First of all, the results can hopefully be checked at a later date. Further- more, most of the really important attempts to do this kind of semo- logical analysis of language have been done by people who are effec- tively bilingual in English and the language described. (cf., Stennes, 1969 and Tab r, 196 ) In such cases it is possible to intuit analy- tical decisions. If meaning can only be analyzed by recourse to in- tuition or upon the acquisition of a very high level of user competency in a language, significantresults in this endeavor will be seriously limited. At least one question posed for this thesis has been, can a technique be devised to analyze the meanings of speech without having internalized all of the principles engendering it? It seems there is no conceptual barrier to this kind of analysis. The preceding analysis of a semological network is presented in evidence. There are obviously some serious weakness in the analysis and some extremely tentative conclusions presented. However, it is safe to assume that these weaknesses and uncertainties could yield to 161 careful application of more data. This is to say, there seems to be no conceptually insurmountable obstacles to accurate analyses of semology using the technique of comparing lower strata emic units in a variety of contexts--a major technique long used in the analysis of grammar and phonology. By the same token, semological analyses must submit to the same limitations imposed upon any other level constructs. Grammatical, phonological or semological analyses can only be judged by their ability to account for and predict actual behavior, that is speech. Thus, any proposed analytical construct is susceptible to refinement and revision a as new behavior is observed and must be accounted for. Recognizing the limits of the construct presented here, there re- mains only to point out what it reveals about the Kapsiki coneeption--and possible perception--of the universe of reality in which he lives. In light of the ethnographic sketch introducing this study, several possible areas of illumination suggest themselves. Before discussing these areas an important precaution concerning the relationship of world view and language must be voiced. A great deal of effort has been expended to demonstrate that langzage reflects, or does not reflect, world view. Usually the evidences adduced are correlations, or lack of corvclations between the language similarities or differences and socio-cultural similarities and differences (cf., for example, Bright and Bright, 1965). One of the serious problems encountered in putting hypotheses to this kind of testing is that, whereas it is generally recognized that cm: the language side, speech behavior does not reflect a direct, isomorphic relationship to the world of reality it transmits informa- tion about and symbolizes, but rather represents a complex process of transduction, the same is not recognized or at least the process is poorly understood, on the socio-cultural side. It seems reasonable to assume that if cultural and social behavior also reflects a world view, this too can only be understood when the 'realization rules' (that is, internalized principles or culture) engendering this behavior have been analyzed. Since presumably the abstract processes by which the inculcated conceptions concerning the real world is transduced to symbolic expression 162 as cultural behavior could include the processes of neutralization and portmanteau realizations, similar world views could result in very dif- ferent kinds of socio-cultural organization as well as very different languages. This kind of analysis of the socio-cultural organization has not been done for the Kapsiki. For this reason the following areas of possible relationship between the language and a particular view of reality must assume the status of suggestions for further investigation. However, the foregoing analysis of the semological structure would 5 strongly suggest that such investigations would be fruitful. a. 5.h.l A Dynamic View of Reality. From a grammatical point of view, most grammatical constructions can be viewed as verb-centric. At the semological level, it seems most i " fruitful to analyze most event configurations as action centered and to explain other phenomena in terms of their relationship to the action event. It has also been noticed that statistically the active forms of verbs far out number the stative forms. Furthermore, it has been point- ed out that many relationships expressed in English by adjectives, per- haps indicating that qualities appear to be viewed as inherent (and thus symbolized on a higher level in the sememic stratum) rather than states resulting from an action, as may be the case in Kapsiki. In light of these facts, the question poses itself, 'does all of this reflect a world view that is basically different from ours, for ex- ample, in that theirs is in some sense more dynamic?’ The answer is not easily forthcoming. The kinds of non-linguistic behavior recorded in my corpus do little to support such a view. 5.h.2 Spatial Dimensions of Actions. In the ethnographic sketch mention was made at several points of the relationship between the people and the land. They are mountain people and their entire pattern of living reflects this fact. The hills provide a feeling of security, they are the focus of religious ritual and mythology and their houses are oriented to the lay of the land. In short this feature of the physical environment constitutes an important element in the universe experienced by the Kapsiki. Few actions can take place which are not affected by this. 163 It is not surprising that verbs symbolizing the actions of 'coming' and 'going' should reflect this orientation. However, a couple of significant theoretical questions pose themselves in this respect. First how did the verbs acquire this characteristic and secondly, how and why did class I verb stem formative suffixes acquire the meaning of Spatiality? These questions involve the problem of language and its relationship to the physical world as well as the problem of language change. It has been suggested that the semology, standing as it does juxtaposed to the world of experience, should re- flect most sensitively the phenomena encountered here. The concep- tualization of the spatial contexts of action as semological aspect would seem to bear this out. Although stem formative suffixes are not rare in African languages, for the most part they indicate dimen- sions of action quite different from those represented by those in Kapsiki. For example, in Fulfulde, Stennes lists among the meaning of the extension suffixes, repetitive or reversive, associative or in- tensive, causitive, manner, reciprocal and dative with one distal suf- fixe, indicating, "that the action of the root is taking place at a distance from the speaker." (Stennes, 1967. 152-155). Some Kapsiki suffixes too, as has been discussed, do not have a spatial referrent and others seem to be used with two quite different (although perhaps logically related) meanings. For example, the suf- fixe 2239 can mean "do something completely or finally". On the other hand, it can carry the notion of seperation or removing away in a spa- tial sense. Etymologically it appears related to the verb BEE "die" or "death". Notice the following usages of :mtg and Egg: kwa mté In the bush. (PSh) in dead xwa mté In enemy villages. among dead ntigfi xe kadzemté They went away. left they go-away kwanaxémte tlené g'yé... After we finished all of the finish-all work we work... (SS2) naxe kekelemte wusu me (She) took the thing away from finish took-from thing mouth the mouth. (P01) '5 keséxwémte xwa (He) wiped the knife dry. (wS3u9) wiped-all knife 16k 'a kebelemté sugwfi He broke the stick. (WSBYS) broke-entirely stick dzemté jaxaggalé za a gké Her husband is on a trip. (PSI) go-away trip man of her The above examples would suggest that the suffixe, depending on the action symbolized by the event and realized by the verb. repre- sents a sense of finality and/or removals (The same kind of observation can be made with the suffixe :yi_which with class I verbs represents a sense of downward motion or of doing something with the intention of L leaving it.) Whether this kind of dual meaning reflects a process of meaning shift in the direction of Spatiality is a moot question. It seems feasable to suggest that when the Kapsiki entered the g" mountain area they used the suffixes on class II verbs to reflect the actions of coming and going over mountainous terrain. Then by analogy some of the class I verb suffixes, having the same phonemic shape, came to express spatial dimensions of action roughly similar to that expressed by their class II counterparts. That the process is not completed or that the original meanings still have significant functions could account for the two meanings attached to some of the suffixes. This reconstruction, of course, represents only one of the pos- sible processes explaining the present state of events. If, however, it should prove correct, it would serve to illustrate something of the kind of relationship existing at the sememic-reality interface of language as well as the process of language change occasioned by changes in the world of reality. A partial check could be made by searching for cognates to the verb suffixes in related languages, particularly in areas near the re- puted origins of the Kapsiki, especially since these areas are consid- erably less mountainous than that now occupied. If cognates could be found which didn't reflect a spatial referent this would help confirm the above hypothesis. At any rate the presence of stem formatives, realizing semolo- gical phenomena more basically related to the event than temporal or modal dimensions of action and which have a spatial referent in the world of experience outside the language, suggests the kind of 165 preoccupation with the mountainous features of the environment which is reflected in other aspects of Kapsiki life. 5.h.5 Temporality and Spatiality. Although expressions of temporality and Spatiality occasionally overlap in English, the notions of time and space are ordinarily kept conceptually distinct. This does not appear to be the case for the Kapsiki. Several linguistic evidences reflect this. In the first case, the verb fig "come" is, as has been discussed, used to mark a temporal context of action in the realization of action events. There is also some tendency to use I; "go" as a marker of future action. This, of course, closely parallels the English and French usage of 'go' to express future action. In addition to these, the preposition xwa meaning "among" when used as a locative marker also fUnctions as a temporal marker to mean "during". For example, xwa va ndla p6 cs "During the rain the house fell." (Gr69) Furthermore, two of the stem formative suffixes discussed above, :39 and :53 are often used with no discernable meaning other than to mark an action as completed. There is also in the language two prepositions taking the same phonemic shape which are ordinarily used as locatives. It seems possible that :te_and.tg_as well as :kg and kg are cognates serving as temporal-spatial markers. SECTION FOUR CONCLUSION The goal of this thesis has been an examination of the use of verbs by speakers of Kapsiki. This attempt represents more than a straightforward linguistic analysis of the verb system. Starting from speech, the vocal behavior engendered by the language, the verb was analyzed, first as a grammatical entity, that is, as a part of the language and then as time symbolic representation of an aspect of the universe experienced by the Kapsiki. Underlying this analysis and accounting for the particular for- mat used in the presentation, were several basic assumptions. First of all, the assumption that language is an abstract set of principles learned by members of society by which they encode information concern- ing their perceptions of the world about them into vocal symbols. Furthermore, this same set of principles is used to decode sensorially perceived vocal symbols resulting in the conceptualization of the world of reality approximately as perceived by the speaker. A second assumption is that the stratificational model (cf.,2.0 for a discussion of this) serves to adequately explicate the language and to explain the speech data. Finally, I have assumed that every society, that is, group of indiviuals behaving by a common set of principles or culture, including the use of a common language, conceives of and probably perceives of the world of experience in a particular and unique way. The language reflects these cognitive and perceptual processes. Part of the goal of this presentation, therefore, has been to specify how the Kapsiki conceives that segment of the world of reality symbolized in the language by the verb system. The methodology involved primarily an examination and comparison of verbs in various grammatical constructions and then a comparison of the various "segments" of reality apparently symbolized by these construction. 166 167 This resulted in the postulation of a number of sememes (lin- guistic units enjoying an isomorphic relationship to perceived segments of the world of reality) and to the formulation of a number of "rules" serving to realize these units into grammatical units. From this it has been possible to make the suggestions included in the preceding section (5.h) concerning the Kapsiki conceptualization of reality. ‘While the evidence is far from unequivocal, it would appear that the Kapsiki view their world in a manner quite different from others (including our own -- although the exact nature of our own world view is impossible to specify.) This, of course, is in keeping with the findings of other attempts to study perceptual differences cross-culturally. For example, Segal and Campbell, after examining the responses to geometric illusions by individuals from 15 societies, concluded, that to a substantial extent we learn to perceive; and that in spite of the phenomenally absolute character of our per- ceptions. they are determined by perceptual inference habits... (Segal, Campbell and Herskovits, 1966. 21k) For example, the discovery that the spatial context of action appears to be conceived as so intimately associated with action as to be an important and frequently symbolized dimension of that action, (cf., 5.h.2) tends to support another conclusion of Segal and Campbell, that perceptual differences may be produced ecologically. (cf., Ibid, 185). This contention seems to be born out by other investigations, notably that undertaken by Allport and Pettigrew and reported in their article, "Cultural Influence on the Perception of MOvement, the Trapezoidal Illusion among Zulus", (Journal of Abnormal Social Psycholog, 1957). In this it was reported that Zulus living in rural areas responded differently from urban dwelling Zulus to the illusion created by a rotating trapezoidal window. In the introduction to the language section of this thesis (2.0) several general theoretical problems were raised and the sugges- tion was made that perhaps this analysis would provide a step toward finding a solution to them. First, the general claims of the stratificational model to ac- curately reflect the relative position of language vis-a-vis the 168 speaker's world of reality appears to have been vindicated. Secondly while the discussion of semology has shed some light on the nature of the language-experience relationship, little has been discovered of an unequivocal nature concerning the causal effects the two exert on each other. However, in the discovery of an imputed spatial dimension to actions symbolized by class I verbs, one possible example of this mutual effect is suggested. Finally, while once again the intimate relationship between the structure of a language and the segmentation and systemitization of the universe of experience symbolized by speech behavior has been demonstrated, no positive evidence has been adduced to indicate whether we are, in fact, dealing with purely cognitive or with cogni- tive and perceptual phenomena. In summary, the thesis presents a description of the Kapsiki culture, social structure and language. From the analysis of the language and the description of the socio-cultural context within which the individual acts, some of the meanings expressed by the verbs have been analyzed. This in turn has led to the formulation of several possible hypotheses concerning the world view of the Kapsiki and of the relationship between this perception of reality and the nature of the language. The first sections are presented as an attempt to add to the scholarly knowledge of a little researched and little understood part of the world, the Chadic speaking people of Africa. The latter sec- tions are presented as groundwork for future empirical research into the area of the relationship between language and non-linguistic be- havior, viz, human perceptual and cognitive processes. BIBLIOGRAPHY BIBLIOGRAPHY Allport, G. W. and T. F. Pettigrew 1957 "Cultural Influence on the Perception of Movement: the Trapezoidal Illusion among Zulus," in Journal of Abnormal Social Psychology. 55, th-ll5. Bright, Jane O. and William Bright 1965 "Semantic Structures in Northwestern California and the Sapir-Whorf Hypothesis," in Formal Semantic Analysis, special publication of the American AnthrOpologist. 67, No. 5, part 2, 2u9—258. Brown, Roger 1958 Words and Things. The Free Press, Glencoe. Brown, Roger and E. H. Lenneberg 195h "A Study in Language and Cognition," in Journal of Ab— normal Social Psychology. 59, ash-A62. ‘ Carroll, John B. 1956 Language Thought and Reality: Selected Writings of Benjamin Lee Whorf. The MIT Press, Cambridge, Mass. Gleason, H. A. 1961 An Introduction to Descriptive Linguistics. Holt, Rine- hart and Winston, New York. ' 1964 "The Organization of Language: A Stratificational View," in Linguistics and Language Study. Stewart, William, Ed. (Report of the Fifteenth Annual Round Table Meeting on Lin- guistics and Language Studies.) Georgetown University Press, Washington, D. C. Greenberg, Joseph 1966 The Languages of Africa. Indiana University, Bloomington. 1968 Anthropological Linguistics: An Introduction, Random House, New York. Hoenigswald, Henry M. 1960 Language Change and Linguistic Reconstruction. The Uni- versity of Chicago Press, Chicago, Ill. Hoffman, Carl 1965 A Grammar of the Margi Languagg. Oxford University Press, London. 1966 Simple and Extended Verb Stems in Kapsiki. Mimeographed paper presented to the West African Language Conference, Yaoundé, Cameroun. 170 Hockett, Charles 1958 A Course in Modern Linguistics. Macmillan, New York Ladefoged, Peter 196A A Phonetic Study of West African Languages. Cambridge University Press, London. Lamb, Sidney M. 1966 Outline of Stratificational Grammar. Georgetown Uni- versity Press, Washington. 1965 "Kinship Terminology and Linguistic Structure," in Formal Semantic Analysis. 0). Cit., 57-64. Lambezat, Bertrand Les Populations Paiennes du Nord-Cameroun et de 1'Adamaoua. Presses Universitaires de France, Paris. Padleweski, André M. 1960 Etude démographiqne de trois ethnies paiennes du Nord- Cameroun: Matakam, Kapsiki, Goudé. Institut de Recherches Scientifiques du Cameroun, Yaounde. Samarin, William J. 1965 A Grammar of Sango. Hartford Seminary Foundation, Hart- ford, Connecticut. Segal, Marshal H., Donald T. Cmpbell and Melville J. Herskovitz 1966 The Influence of Culture on Visual Perception. Bobbs- Merril Co., New York. Stennes, Leslie H. 1967 A Reference Grammar of Adamawa Fulani, Michigan State University, East Lansing. 1969 The Identification of Participants in Adamawa Fulani. Unpublished Ph.D. Thesis, Hartford Seminary Foundation, Hartford, Conn. Taber, Charles R. 1966 The Structure of Sango Narrative. Hartford Seminary Foundation, Hartford, Conn. Williamson, Kay 1965 A Grammar of the Kolokumo Dialect of Ijo. Cambridge University Press, London. APPENDIX APPENDIX THE CORPUS The corpus on which this analysis is based is very ruoghly es- timated to consist of about 75,000 words collected by me personally during my stay in Mogode, Cameroun between March 1967 and April 1968. For the most part the material was first taped on a Uher hOOO Report-L tape recorder then, with the help of my assistant Mr. Luc Sunu or an- other informant, transcribed and glossed. A number of speaking styles and a wide sample of speakers is included in the corpus. The examples given, with the exception of a few short ones, and indexed to indicate who the informant was and what style of speech they represent. The following set of notations has been used to identify the examples in the text. These symbols appear in parentheses at the ex- treme right of the page. The first letter indicates the source in my notes and for the most part reflects the style of the speech from which it was taken. The second letter refers to the informant and the number indicates the number of the discourse in my notes. Thus, on page 57, line 26 the notation (POA) means, "picture elicitation, André Ousman, picture number A". Following is a list of the abbreviations used in the thesis. DISCOURSES P Picture elicitation--discussions describing pictures of com— on activities. C Conversations between two or more people. N Narratives--stories of ordinary events by individuals. S Short elicitations, except where followed by 'B'. SB refers to sermons by Martin Bera W Word, phrase and sentence lists. T Texts--recitations of common and current events (cf., N above). 171 172 F Fables, stories of old, myths, etc. Gr A collection of miscellaneous material including sentences, utterances elicited while checking and specific elicitations, using a number of different informants. INFORMANTS A The Aranado or village chief with the name, Maze wusaxwaxwele. No French, some pidginized Fulfulde. B Martin Bera, Catechist for the protestant church. A native of Guria. Sermons taken from a record. Some French and fluent in Fulfulde. C Timotné Teri Pouh, Catechist for the protestant church in Mo- gode. Native of Mogode. Some French fluent Fulfulde. J Joseph Zera Mpa, has a few years in French school at Siir, French and Fulfulde quite good. Young man (16?). K Kwade De, young girl, no French of Fulfulde, native of Mogode. M Marta (wife of Tlimu, see below). Some French and Fulfulde education, speaks both well. 0 Andre Ousman, student at the French elementary school at Siir. French fairly good, Fulfulde good. P Pierre Vandu, school teacher at Siir. Native of Mogode living in Siir since 1961. French good, Fulfulde fair. S Luc Sunu, young man (20?) native of Mogode, several years of formal French, French and Fulfulde both good. T Yakubu Tlimu, native of Mogode, studying to be protestant pastor, some French training in mission school, French poor but fluent in pidginized Fulani. Z Matthieu Kweji Zama, young man (18?), native of Mogode, no French schooling, French poor Fulfulde fairly good. .l, .. it“s“..flfi. t 1! O :n‘i I) .r.‘ a .1"! u u I‘lflv I .iv {Lilo-alt i R "IIIIIIIIIIIIIII’II'IWA” III“