The Canela (Eastern Timbira), I: An Ethnographic Introduction.
By William H.Crocker
Smithsonian Contributions to Anthropology,
Number 33, 487 pages, 11 tables, 51 figures, 78 plates, 1990.

<<back table of contents  next>>

<<return to literature


This monograph, which will be referred to as “Canela Introduction” for brevity, constitutes an overview of my ethnological field research among the Canela carried out between 1957 and 1979. As such, it is meant to serve as an ethnographic handbook for general ethnologists throughout the world wherever Smithsonian Contributions to Anthropology is sent and exchanged. One purpose of this work is to reconstruct and relate what has happened to the Canela since the time of Curt Nimuendajú, who studied them between 1929 and 1936, and wrote his masterpiece, “The Eastern Timbira” (1946), on this tribe. Another purpose is to serve as an introduction for a series of volumes [Pr.1] to be written during the next two decades or more. As an ethnographic handbook, this monograph is primarily descriptive and is focused in places on the presentation of data-rich field materials. (For the explanation of the bracketed codes and for their interpretation and use, see [In.4.a].)

[In.1] The Canela

The Canela (more correctly, the Ramkokamekra-Canela) of the state of Maranhão, Brazil, live just east of the watershed of the Amazon basin and just west of the drought-ridden Northeast. They are a tribe of South American, Northern Gê-speaking Indians, like two better known tribes, the Kayapó, who occupy areas to the west of the Araguáia River, and the Apinayé, who live near it (Map 1). The Canela probably experienced their first indirect European influences in the early part of the 18th century, or earlier [II.A.3.a.(1)]. They survived, however, to make peace with local pioneer authorities in 1814 [II.B.1.a]. They surely numbered between 1000 and 2000, aboriginally, but were about 300 in the 1930s, 400 in the 1960s, and 600 by 1979 [II.A.2]. The Canela combined hunting and gathering in “closed” savanna lands (cerrado) [II.A.3.b.(2)] with some horticulture, but in the 20th century they have had to rely on slash-and-burn techniques for cultivating bitter manioc and dry rice, currently their principal crops [II.C.3].

The Canela were a comparatively inward-looking society aboriginally, being largely endogamous and relatively self-sufficient economically [IV.C.1.f.(1)]. They relied little on trade with other tribes. By the 1970s they were dependent on the federal National Indian Foundation (Fundação Nacional do Índio, FUNAI) (Glossary) for protection and medicine, and on local interior farmers (“backlanders”) for economic support through sharecropping and through rendering simple services for food [II.C.3.g]. By the mid-1970s, their economic viability and group morale had improved considerably in contrast to what they were in the late 1950s and 1960s [II.B.2.i]. The population expansion of one-third between 1969 and 1979 was due largely to their more confident outlook and to better medicine [II.B.3.i]. Both of these improvements were due to the greater care and concern of the National Indian Foundation [II.B.2.i.(2)].

The Canela have survived as an independent tribe speaking its own language because their ancestral lands have had few economic assets for Brazilian nationals to exploit and because no thoroughfares, either rivers or roads, pass through their immediate area [II.A.3.c] (W. Crocker, 1964b). (For traffic routes around the Canela area, see Map 2 [II.A.3.a.(2)], and for continuing overview material on the Canela, see “Outstanding Ethnology” [I.A.1]. Nimuendajú’s (1946) work on the Canela is entered in the Human Relations Area Files (New Haven) and noted as “Ramcocamekra” by Murdock (1967) in his Ethnographic Atlas. The Canela in Price (1989) are listed under the entries “Ramko-Kamekra,” and “Canela.”)

[In.2] The Apanyekra

The Apanyekra-Canela tribe, which will be referred to simply as the Apanyekra, is culturally so similar to the Canela that it is difficult for outsiders to find significant differences between them. The Apanyekra numbered about 175 in the late 1950s and increased to around 250 by the mid-1970s. They live between 45 and 55 kilometers to the west of the Canela (Map 3), depending on the varying locations of the two tribal villages within their own lands. These two Canela tribes speak the same language of the Gê family as the Krahó, who live some 300 kilometers to the southwest (Map 1).

About 15 percent of my field research time was spent with the Apanyekra, and the data from these people are used for comparative purposes with those of the Canela. (For Ramkokamekra-Apanyekra ecological contrasts, see [II.A.3.d].)

[In.3] Objectives and Structure of This Volume

The objective of this introductory volume is to give the reader an overview of the Canela world and its operation. Several topics ([II.B] [III.B.2.f] [III.E] [IV.A,B,D] [V.A]) already noted in the Preface [Pr.1] are intended for more

MAP 1.―Gê-speaking Indians of Central Brazil (roman capital letters) and their neighbors (in italics), and Timbira tribes of the 1970s (underscored) (rivers in smaller type).

comprehensive treatment in future monographs. This monograph includes as full treatment as the data so far collected allows on the following subjects: social groups [III.C], chieftainship [III.D.1], council of elders [III.D.2], judicial system [III.D.3], and marriage [III.F]. There is a third category of topics on which relatively little material was collected in the field: linguistics [Ap.4], geography, specific environmental studies [Pr.1], [II.A.3.d], [II.B.2.g], general ecology [II.A.3], and material culture [II.G]. Other topics that could be significantly expanded are: three of the annual cycles (climatic, environmental, and economic) [II.C.1,2,3], three aspects of recreation (sports, games, and body painting) [II.F.1,2,5], the daily cycle [II.E], socialization [III.A], and psychological aspects [III.B].


This monograph consists of five parts, an epilogue, and six appendices (see “Reference Strategy” [In.4.a]). Part I: “The Field Situation” provides a view of the fieldwork practices and exposures.

The first chapter [I.A], “General Characteristics” starts with a description of the tribe to provide a background for understanding my experiences while living among the Canela. This is followed by a report on my daily activities, as a structure for presenting some typical field involvements and experiences.

“Early Acceptance Experiences” [I.B] focuses on the more obvious and most difficult field adaptation problems, and as such should especially interest students.

“Problem-solving in the Field” [I.C] presents the relative advantages and disadvantages of the fieldwork conditions, which were more pleasant than with most tribes. However, the relatively large populations of the Canela and Apanyekra, their deficit economies with their moderate endemic hunger, and their contact of over 100 years with coastal cities and their resulting awareness of city prices made fieldwork more difficult than with many other Brazilian tribes. Moreover, their tradition for “begging,” their aggressive attitude against property retention, and their great expectations of the ethnologist because of Nimuendajú’s past largess created further difficulties.

“Field Equipment” [I.D] provides information of obvious utility to other fieldworkers. This chapter furnishes a history of what was technologically available to ethnologists over a period of more than two decades.

“Learning the Canela Language” [I.E] is a personal narrative, because learning the language was my special pleasure and pastime and of great advantage in establishing rapport and communication with the Canela. In contrast, it had not been undertaken seriously by other Timbira ethnologists. Moreover, I had the collaboration of a SIL missionary-linguist in the 1970s. The more technical linguistic information is in Appendix 4.

“Diaries and Tapes” [I.F] represents an enormous investment of my personal time and energy but supplies a wealth of material for future analysis [Ap.3.a].

“Special Research Assistants” [I.G] is the most complete account of my fieldwork techniques, and should be of interest to scholars and students.

“Special Friends in the State of Maranhão” [I.H] addresses the difficulty of thanking so many field colleagues with whom I collaborated over a period of 22 years. In this chapter I am acknowledging their help, as well as describing the field situation. Of course, many individuals to whom I am indebted have had to be left out of even this expansion of the acknowledgment section, but my principal benefactors in Maranhão are included here, and I hope they feel my deep gratitude both for their help and for their friendship.


This part presents ethnographic data that are intrinsically important, but which are not the principal focus of this monograph. Thus, these materials are presented here in what is primarily a descriptive manner to provide background information.

These background materials are arranged in three phases: (1) ecological context-oriented: the linguistic, demographic, geographic, historical, and acculturative contexts; (2) natural and cultural cycle-organized: the climatic, environmental, economic, ceremonial, life, and daily cycles, and (3) expressive and material culture-focused: recreational, musical, athletic, body decorative, and material culture.

The first portion of the first chapter [II.A], deals with general ecology, and the position of the Canela among their related language families―the Eastern Timbira, Timbira, and Northern, Central, and Southern Gê (Maps 1 and 4). After a small section on demography (Gê population numbers), the rest of the chapter deals with the environmental and social contexts of the tribe.

The second chapter, “Diachronic Context,” [II.B] is crucial for understanding the Canela’s present condition, as seen by examining its past. The first section, “Indigenous Accounts,” starts with accounts of the earliest tribal contacts. It is followed by pacification in 1814 and ends with the time of Nimuendajú in 1929. These events continue the sequence of mythological stages recorded in part of “Oral History and Cosmology”


The second section, “Acculturation,” continues the narrative from 1929 through 1979 with historical facts and data collected in the field. (The Epilogue continues certain threads of the narrative through 1989.) While relatively long, this chapter is still only a summary account of the available materials and constitutes overview coverage for the planned monograph on changes and conservatism [Pr.1][II.B].

The final section is a limited history of the município (city and county: township) of Barra do Corda in which both tribes live. It is included here to explain a principal part of the external sociocultural context of the Canela.

The third chapter [II.C] is on the four annual cycles―climatic, environmental, economic, and ceremonial. It furnishes several of the critical contexts for understanding the timing of events, and their maintenance of the Canela sociocultural system. For the climatic cycle (Table 1), the data are limited. The environmental cycle (Table 2) is similarly limited. Additional and more developed and extensive material can be found in “The Eastern Timbira” (Nimuendajú, 1946:1–2,37–76). The economic cycle is presented in a more complete manner (Table 3) and is closely related to certain important acculturation problems. The ceremonial cycle constitutes an overview of the chapter on festivals [IV.A] and is presented at this early stage to provide an appreciation of the great scope and nature of the festival system.

The fourth chapter [II.D] focuses on female and male life cycles, a further narrowing of the scope of the background materials being presented: (1) from the general to the particular, (2) from what is external to the tribe to what is internal, and (3) from the impersonal forces of ecology and history to a closer study of the female and male individuals through depicting their cultural cycles. This chapter also includes a description of the rites of passage(Table 5). Thus “Life Cycles” serves as a preparation for understanding and appreciating subsequent discussions regarding socialization, social and ceremonial units, marriage, tribal festivals, and individual rites.

“Daily Cycle,” the fifth chapter [II.E], presents the ongoing events (daily activities (Table 7) and expressive culture) of the Canela way of life. This chapter includes the various divisions of the day (Table 6) and how the Canela refer to past and future days (Figure 12). It also provides information on the three daily social dancing sessions, on the work period, and on the afternoon log races and track events, as well as on the two daily council meetings. The relative importance given to work, pleasure, and social relationships becomes evident through this material. It is here that the Canela may come alive for the reader.

The sixth chapter, “recreation” [II.F], furnishes more material on daily activities and additional material on expressive culture. It includes music, sports, games, and body painting. Choral singing (their primary musical form) plays a large role as a daily recreational event. Singing is done in the context of a festival only about one tenth of the time, and even then it is usually more secular than sacred. Extensive choral and individual singing are recorded on tape [Ap.2.d.(1)] and would be useful to ethnomusicologists [Ap.3.e]. The section, “Sports,” includes log racing, certain track events, and archery. Children’s games, as a category outside of sports, are quite limited. Body painting occurs as a daily practice far more often than as a ceremonial one.

The last chapter [II.G], is concerned with the material artifact collection (Table 8), which is quite extensive [Ap.2.a] [Ap.3.g]. Although many individual artifacts are both ceremonial and used on a daily basis, it is convenient to deal with them in one chapter.


The first chapter [III.A], a study on socialization, was carried out in the late 1950s (including 1960). A contrasting view of socialization in the 1970s is provided, as well as significant changes affecting socialization that took place much earlier in the 20th century as a result of culture contact. At the end of the chapter, the focus is on ethno-ideology and the growth stages through which (Table 9) the individual passes. No monograph on socialization is planned in the 8-volume series, although a study and analysis of the 120,000 feet of 16 mm film [Ap.3.h]1 would be useful in a study of this subject.

The second chapter [III.B], on psychological aspects and behavioral orientations, follows the chapter on socialization. Adult behavioral matters are partly a result of specific kinds of socialization, though the relationship constitutes a two-way process. These Canela value-laden orientations and behaviors are presented as polarities, such as individuality within solidarity, the “little good” with the “little evil,” generosity versus stinginess, and the inner against the outer or “we” against “they.” The abundant data from the manuscript and tape diaries [Ap.3.a], and the 16mm film sound-track translations of conversations on socialization [Ap.3.d], constitute significant additional materials not elaborated in this monograph.

The third chapter [III.C] delineates, differentiates, and defines almost all the various social and ceremonial groups (e.g., moieties, age-sets, men’s societies, rituals owned by families) that comprise the Canela sociocultural system and make social and ceremonial life so complex. Again, no attempt has been made to separate what is social from what is ceremonial. The various socioceremonial units presented here are also discussed in subsequent chapters.

Once the various socioceremonial units are identified and defined, the next step is to discuss how they operate. The fourth chapter [III.D] describes the Canela political system, including the chieftainship, the council of elders, and the judicial system. The operation of these socioceremonial groups demonstrates social organization at the macro level. Politics is the most difficult subject for any researcher to probe without a fluent knowledge of the language, including the ability to collect information through simply overhearing native discussions and debates [III.D.1.g.(1).(c)]. Since personal aggression and the assumption of positions of superiority are considered evil by the Canela, political activities are carried out in a very subdued way. No one must let it appear that he is running for office or behaving in a political manner, that is, he must avoid the appearance of directing and maneuvering other individuals into positions that may or may not be advantageous for them. Consequently, relatively little material was collected on the chieftainship. Most of the data on the council of elders (taped meetings) and the judicial system (taped hearings) are on tape in the Canela language, still untranslated [Ap.3.c]. Therefore, their exposition will have to wait until a great deal of time has been spent on translating the tapes [Pr.1] [Ap.2.d.(2)] [Ap.3.b]. Because none of the planned monographs will be on the political system, as much material as possible is included here.

Following the macro elements of social organization come the micro ones: kinship (Table 10) and other relationship systems [III.E], and marriage [III.F], the more informal and personal social structures. Extensive data have been collected, both qualitative and quantitative, for the kinship and other relationship systems. Some of these data have already been published (W. Crocker, 1977, 1979). This chapter includes the following sections on terminological relationship systems: consanguineal (Table 11), affinal (Figure 28), name-set transmission, Formal Friendship, Informal Friendship, mortuary, teknonymy, co-father, and ceremonial.

Data on marriage were previously published (W. Crocker, 1984a:63–98). The same and additional data are supplied here so that the materials in this chapter are nearly exhaustive, but also include a number of illustrations on “blood” ethno-ideology (Figures 38–44).


In the broadest sense, all the material here is related in some way to religion. Ceremonialism (tribal festivals and individual rites) constitutes the dramatic and behavioral side. In contrast, belief systems comprise religion’s more ideational side: other-time-projected myths (oral history), other-place-projected myths (cosmology), and directly practiced concepts (shamanism, pollutions, shamanic practices).

The “Festival System” [IV.A] is the macro element in the sector of ceremonialism, and “Individual Rites” [IV.B] is the complementary micro element. Festivals (Table 4) are held for the entire society, while individual rites are staged for just one person and are put on by her or his kindred. There are two kinds of festivals: the annual cycle festivals and the great dry season Wè?tè festivals, one of which is put on every year.

There are two foci in “Individual Rites”; one is on the rites themselves, rather than on the roles individuals play in them, which is stressed in the chapter “Life Cycles” [II.D]. Thus, while “Life Cycles” gives a quick portrayal of materials to provide orientation and comprehension for following chapters, “Individual Rites” furnishes the descriptive detail needed for comparative studies. The second focus is on the contribution these rites make to reinforcing the sociocultural context. These individual rites follow each other sequentially from birth through death. They do not, however, constitute all the “rites of passage,” taken in the traditional sense, encountered by the individual as she or he grows old. Some of the tribal festivals, such as the Khêêtúwayê and Pepyê, must be considered traditional rites of passage, though they are put on by the tribe rather than by the individual’s kindred. Some individual rites are related to marriage and so are better described as part of a sequence in “Marriage.”

There are two chapters on the ideational aspects of religion. Other-dimension-projected beliefs, “Oral History and Cosmology,” [IV.C] comes first, as it is large in scale; while the chapter on directly practiced beliefs, “Shamanism, Pollution, Medicine, Affirmations, Transformations” [IV.D], follows, as it is smaller in scale. The oral history (other-time-projected) and cosmology (other-place-projected) chapter is also divided into these two parts. Rather than being based on the entire collection of myths and war stories, it is a native ethnohistory reconstructed by Canela research assistants themselves. The sequence starts at the Canela beginning with the stories of Sun and Moon, passes through the cosmology and festival “gathering” period of tribal travels, and ends with the pacification of the Canela. The continuation of this “history” of the Canela is found in the section on indigenous accounts [II.B.1].

The directly practiced beliefs found in the fourth chapter are related to shamanism, witchcraft, ghosts, pollutions, food and sex restrictions, medicines, ethnobiology, psychic abilities, self-transformations, breathing strength into substances or people, and gaining strength through singing “strong” [IV.D] words. Materials collected on these topics are extensive and are closely related to individual rites. The eventual plan is to publish these two kinds of materials in one monograph [Pr.1] [IV.B.D].


This part starts with the “native’s view” approach, a kind of “ethnoscientific structuralism,” which is pervasive throughout most of the various Canela sociocultural sectors. The various kinds of dualities and triads identified are categorized on the basis of Canela expressions that may be related to their thought patterns.

The two chapters of Part V come partly from a previously published article (W. Crocker, 1983). The plan is to present this material [V.B] in greater depth later [Pr.1].

This structuralism is ethnoscientific, having evolved through extensive work with research assistants in the field and representing what they see. Inarticulatable traditional patterns were made articulate and then were applied to, or found in, many sociocultural sectors. The resulting Canela “dualism” (Glossary) is both complementary and oppositional. Culturally paired items are seen to be in either a facilitating or in an oppositional relationship with each other. Combinations of these paired items constitute triads, some of which are fixed, while others are modifying, and still others are generating. It is through the amelioration of the stresses in transformational triads that the Canela see some cultural problems as being mediated or resolved.

This structuring, or pairing, is supported by specific Canela expressions: aypën katê, two items related as a pair, in either a complementary or an oppositional manner; ipiprol, two paired items related in complementarity (in parallel); aypën kunãã mã, two paired items related in opposition, or, in this case, standing opposite and facing each other. This approach may constitute a step toward a different kind of structuralism.


The missionary-linguist Jack Popjes (Figure 11) and his wife Jo (Summer Institute of Linguistics) arranged for collecting and sending a considerable body of data on the Canela during the 1980s [Pr.3]. The Epilogue consists of information mostly from their materials. Selecting what is reliable and significant is a problem, so I have chosen to stress political and demographic changes.


The first appendix, “Ten Field Trips to the Canela over 22 Years,” describes the many trips to Brazil, the conditions under which they were made, and the general study topics undertaken on each trip.

Appendix 2, “Canela and Apanyekra Collections at the Smithsonian Institution,” is a quantitative list of material artifacts, products of photography, tapes made in the field, and the various kinds of data written on paper.

Appendix 3, “Primary Materials for Future Studies,” is a descriptive report on the Canela primary data collections, such as the material artifacts, the diaries and tapes, and the various photographic and tape collections.

Appendix 4, “Linguistic Notes,” has linguistic materials that were too complicated to include in the Linguistic Key or “Learning the Canela Language” [I.E].

Appendix 5, “Concept of ‘Today,’” describes the complicated Canela concept of their 36–hour day.

Appendix 6, “Sources of Data,” describes the derivation of data from such sources as daily notes and observations, research assistant council meetings, backland community visits, Indian service agents, library sources, researcher’s memory.

[In.4] Definitions and Editorial Decisions


Rather than constructing the traditional index of names and subjects, and to avoid the technological publishing difficulties of supplying a myriad of page references within the text, a reference system has been devised to help the reader locate desired material within this monograph: the Reference Outline, which reflects all levels of headings by alphanumeric designations. The scope of the Glossary has also been enlarged from a list of definitions to an additional reference system by supplying the appropriate reference codes.

Context references in the Glossary and cross-references in the text are accomplished by providing, in brackets, the alphanumeric designation of the heading for the pertinent material. The alphanumeric designations precede the headings within the text and also in the hierarchical listing of the headings in the Reference Outline. The order of elements in the code employed in these designations is as follows: A citation of a heading within one of the five main parts of the book begins with a capital Roman numeral (I-V), which identifies the part. Next, the relevant chapter is identified by a capital Roman letter, and a major section of the chapter by an Arabic numeral. Successive subdivisions involve lower case Roman letters, Arabic numerals in parentheses, and lower case Roman letters in parentheses. If the citation is to a heading within one of the supporting sections of the book (Foreword, Preface, Introduction, Epilogue, or Appendices), it begins with a two letter abbreviation: Fo, Pr, In, Ep, or Ap. The largest division of one of these sections is marked with an Arabic numeral. Successively smaller subdivisions are indicated in the same manner as those in chapters in the main part of the text: lower case Roman letters, Arabic numerals in parentheses, and lower case Roman letters in parentheses.

For instance, if I want to refer in a certain place in the text to the discussion of how the Canela were chosen as my tribe for field research, I insert the code [Pr.2]. A reference to how a mother chooses co-fathers for her fetus is made by inserting [II.D.2.h.(l).(a)] in the text. If I want to refer the reader to the discussions on the male roles of both hunting and farming, which have the same code except for the final numerical designation [II.D.3.i.(6),(7)] a comma is used to designate the final, dissimilar parts of otherwise identical designations. Two less similar codes, representing female and male loss of virginity, for instance, are also separated by a comma if placed in the same brackets for comparison: [II.D.2.a,3.c]. If I want to refer specifically to material in one of the higher headings and not to the material in its subheadings, which may be a whole section of a chapter, I put the last number or letter of the code in italics. For instance, a discussion of Canela activities that provide opportunities for emotional outlets to relieve frustra­tions is designated by a chapter heading [II.F], and the most complete information on wet- and dry-headedness is under a section heading [III.C.7].

Because headings provide the key to the internal reference system, they must adequately, but briefly, represent the material they cover. Thus, the headings in the Reference Outline are expanded (after the colon) to indicate the extent of the ideas covered.

Recognizing that frequent insertion of reference codes in the text may interrupt the reading process, I have depended heavily on the utility of the Glossary to reduce the instance of such codes. Any term or concept included in the Glossary is cross-referenced in the text at its first mention or when it is crucial to understanding the discussion at hand. References to material external to the monograph are made in the traditional way: author, year, and page number are placed in the text in parentheses (Da Matta, 1982:64) and the full bibliographic information appears in the Literature Cited.



A distinction must be made between the folk population who live around the Canela Indians in the interior and the townspeople of Barra do Corda [II.d]. Another distinction must be made for highly educated people living in big cities of Brazil (or educated urban foreigners), who happen to visit Barra do Corda or stay with the Canela Indians for some time. The Canela make these three distinctions [III.D.1.c.(3).(a)], and two of them are also made by the Krahó Indians according to J. Melatti (1967:143).

Near where the Canela live, two Portuguese expressions are used by the interior people to contrast themselves with the Indians: cristão (Christian) and civilizado (civilized person). The people of the interior refer to the Canela and Apanyekra as índios (Indians) or cabôcos2 (i.e., people of lesser status). These terms were socially self-elevating for the interior folk and depreciative and condescending for the Indians, though they were accepted by the Canela and the Apanyekra [II.B.3.e]. The terms civilizado and índio are used here primarily in the context of the Awkhêê acculturation myth, where their negative connotations are pertinent [II.B.2.f.(1)] [IV.C.1.b.(6)]. “Indian” is merely a statement of cultural fact (i.e., a major cultural difference) of living in a tribal state. The Canela accept this term.

Similarly, the term “Brazilian” is used as a cultural term. While there are significant regional cultural diversities throughout Brazil and among social classes, these diversities are minor when contrasted with the great cultural differences between Brazilians and tribal Indians. Nimuendajú (1946) used “Neo-Brazilian” to designate the interior populations of Old World origin that replaced the legitimate (legítimo, as the Canela say) Brazilian, i.e., the Indian. “Neo-Brazilian” is not employed in this book because it is not used by Brazilian ethnologists today, and because it has negative connotations.

The term I prefer for the interior, rural, non-urbanized Brazilian folk of the município of Barra do Corda is “backlander” [II.A.4.d.(1)]. This is the term used in the translation of the Brazilian classic, Os Sertões, by Euclides da Cunha (1973). Thus, “backland” or “backlanders” designates the farming or cattle-ranching folk of the interior who live around the Canela and the Apanyekra villages (Map 3) [II.B.4]. The residents of Barra do Corda are referred to as

“residents,” or as “small-city people” or “townspeople” [II.B]. Individuals who live in the large cities of Brazil (including the state capitals, Brasília, and other large urban centers) and educated foreigners are referred to as large-city people. The term “urban” is used for either small or large-city dwellers.

The term “backlander” is particularly appropriate because the rural culture that was typical of the states of Northeastern Brazil, and particularly Ceará, extended into the Barra do Corda area. Culturally, the folk region around Barra do Corda in the late 1950s was more Northeastern than Amazonian. By the late 1970s, such cultural differences were harder to distinguish.


The term ‘informant” carries a pejorative connotation. Consequently, it is preferable and certainly more respectful to use the expression “research assistant.” In any case, when certain individuals work for an ethnologist for a period of two years or more, they are not just informants; they have become research assistants. Rather than discriminate between individuals who answered only an occasional few questions and those who spent hundreds of hours over a number of years facilitating the collection of field data, the expression “research assistant” is used here for every Canela and Apanyekra individual who gave any significant help to this research effort.


The Canela are known to each other by both their Portuguese and Canela names. The Canela and Apanyekra are known to backlanders, Barra do Corda residents, and local Indian service agents almost entirely through their Portuguese names, which often include surnames as well as Christian ones. Thus, using the individual’s Canela name, instead of her or his Portuguese name, constitutes considerable protection of privacy. When I have to make negative comments about individuals to illustrate a point of discussion, they are anonymous. However, Canela and Apanyekra customarily refer to their chiefs of the last century and the first part of the 20th century by their Portuguese names rather than by their Canela ones: Major Delfino instead of Kô?kaypo. This practice is sometimes followed here when these individuals and their immediate descendants are deceased.

There is another aspect of concealment when using Canela names. Sometimes individuals change their public name from one of their name-set names to another. Moreover, there are circumstances under which an individual can assume the possession of a partially or entirely different name-set and use one of the names in the new name-set [III.E.4.f] [IV.A.3.e.(3)]. Thus, just because a certain individual is identified by a particular name in this volume does not guarantee that she or he will be known by the same name 20 years from now. However, their Portuguese names do not change during their lifetimes.


References made to particular “studies” that took place at specific periods between 1957 and 1979 are referred to as studies in “the late 1950s” (which includes 1960 for practical purposes, unless otherwise stated), and “the 1970s” (which includes 1969). For instance, references are made to the socialization study of the late 1950s, the marriage study of 1970, and to the key words and concepts study of 1979. My fieldwork visits are grouped in these ways. It should be noted that these studies do not directly correlate with the inclusive dates of any one of my 10 field trips.


Zoological and botanical names in Portuguese that have no translations in English are not italicized. If the initial letter of a Canela term, such as Tàmhàk, is capitalized, the word is not put in italics. Descriptive terms, such as “the commandant” of the troop of novices, are not capitalized. When, however, such terms as “Upper” and “Lower” age-set moieties are referred to by the capitalized terms “Upper” and “Lower,” it is understood that these words form part of their proper names [III.C.3].


The Canela orthography used here [In.5] [Ap.4.c] is entirely phonemic, except where intended not to be. Careful translations of certain Canela words and expressions into English are very important for comprehending and appreciating the text. Thus, the Canela phrase is followed by the roughly equivalent words in the same order in English, so that each element of the phrase being translated can be easily identified. When the literal translation is unclear, a second freer translation follows (e.g., Wa ite mãã kuran: I past-tense-indicator emu kill: I killed a South American ostrich [a rhea or ema]).

The use of spaces and hyphens between the same Canela terms varies to express different purposes. Terms may be separated for translation but combined without spaces or hyphens in the text. Terms in Table 8 often vary in this way from those in the text to preserve the form of the earlier presentations of lists of artifacts to various institutions.

A text with many terms in the native language hinders comprehension. Consequently, English equivalents are used wherever possible, including translations. This cannot be done for vernacular botanical or zoological terms that do not have English equivalences such as “buriti” and “paca” (a small rodent), so these terms and their descriptions are presented in roman type. Canela words for which no English or Portuguese translations exist and to which reference must be frequently made are retained here in Canela, such as the Wè?tè girl (a role of ceremonial high honor) and the Pró-khãmmã (the festival-governing age-set in the council of elders).


The frequent use of the terms “fathers’ sisters” and “mothers’ brothers” is clumsy. “Aunts” and “uncles” are used here instead, without quotation marks; but it must be understood that mothers’ sisters and fathers’ brothers are not being included in these terms. These latter categories are “mothers” and “fathers” as understood in the Crow or Canela terminological systems.

Moreover, the terms aunt and uncle (Glossary) are used here in a more general sense to include any of the kintypes in the kin categories of tùy and kêt, such as grandparents, great aunts and uncles, and certain cousins. The reciprocals, niece and nephew, are used in the same way. When a precise kintype (Glossary) (Scheffler and Lounsbury, 1971:2) is meant, it is used, such as father’s sister or mother’s brother. When such a term is used in quotation marks its classificatory sense is meant. A mother’s “brother” means any of ego’s mother’s relatives that she calls “brother” but not her uterine brother. (For a discussion of kinship, see [III.E.2,3].)

It is important to note that because persons other than those in the kin category (Glossary) of “uncle” often carry out an uncle’s social or ceremonial role for ego, especially when this individual’s kin category uncles are far away or nonexistent, this term often includes persons not in this kin category at all, such as fathers or “out”-brothers-in-law [III.E.3.a.(2)]. Thus, “uncle” refers to a traditional role.


Most tribes are matrilocal/uxorilocal (Canela and the Gê) or patrilocal/virilocal (most of the northwestern Amazon), though there are other arrangements. It seems appropriate, therefore, that pronoun references to tribes that are distinctly matrilocal/uxorilocal have feminine pronouns placed first and masculine ones afterward (she/he) and tribes that are patrilocal/virilocal have the masculine gender positioned first (he/she).


Anthropologists have maintained close relationships with the Brazilian personnel of the National Foundation of the Indian (FUNAI: Fundação Nacional do Índio) since 1968, and with the Indian Protection Service (SPI: Serviço de Proteção aos Índios) personnel before then. For simplicity, neither of these services is mentioned by name or by acronym except when an unusually positive contribution deserves recognition. The expression “Indian service” is what North Americans easily understand. In order to make the reference more generic, the word “service” is used here with its initial “s” not capitalized.



These phonemes sound approximately the same as in Portuguese. The phonemes u, ô, and o, and their nasalizations, are rounded, but the other phonemes are not.

[2001: Adjusting to the exigencies of computers and website languages, the dieresis is used here as well as the tilde for phonemic vowel nasalization. WHC]


















ä (rare)














These phonemes are back and unrounded. They have no Portuguese, Spanish, English, or French equivalents. For “high,” “closed,” “mid,” “open,” see Pike, 1947:5.



ù high and closed

ü high and closed

è mid and closed


à mid and open (like “puddle” in English but

more open)

ã mid and open (like canto, in Portuguese but more open)


[2002: The rare phonemic nasalized vowel ä is simply the a nasalized. The frequent phonemic nasalized vowel ã is à nasalized.

One difficulty in deriving Canela orthography arises from the fact that Latin has only five vowel symbols, but Canela has 17 phonemic vowels. Another problem comes from the Brazilian dislike of using the “y” symbol as a vowel, as some authors do to provide six vowel symbols. Thus, through the use of diacritics, five vowel symbols have to be made to represent 17 phonemic vowels. Other problems come from the software difficulties of placing a tilde over some vowel symbols to represent their nasalization and the difficulty of placing two diacritics over the same vowel symbol. Linguists use the reverse cedilla as an alternative symbol for nasalization, but, again, such a solution presents software difficulties at this time, so I am using the dieresis.] 


When following a vowel, the semivowels (phonemes as glides) serve to complete syllable length.

w         west, pew

y          yes, coy

The y ranges from the y in “yes” (most frequent), through the “n” in “new” (Yõ?hê: male name), to the unvoiced “s” in “sky” (-khyê: sibling), and is always palatalized.


Unlike in English, the phonemes /p/, /t/, and /k/ are unaspirated, while /kh/ is aspirated.




p [p,b]

pai, tabu


t [t,d]

tatu, idéia


k [k,g]

capa, igarapé





? [glottal]



ts [4,]

tia (northern Port.)

cents, church






r [r,l]

orar, Isabel (Spanish)












There are no diphthongs in Canela, but vowel length is phonemic: single vowel length (katswa: night) is approximately doubled by the same following vowel (kaatswa: salt). The glides (w and y) complete syllable length, standing in the place of the second vowel (Kawkhre: male name; -impey-ti: beautiful/good-very). Consonants also complete this same syllable length (Pa?päm: God; Kaapêlk: male name; /hapak/ [happak]: ear), which is roughly equal to double vowel length.


Stress almost always falls on the last syllable of isolated words but is not phonemic. Stress is altered and determined at the phrase level rather than at the word level. The acute accent ( ́ ) over a vowel will only be used to indicate stress, whenever necessary, or both the superlative and stress.


All words are written phonemically except when they are intentionally written nonphonemically in order to indicate their most frequent pronunciation. In these cases, the acute accent is added to indicate stress. For instance, the word “buriti” (a palm), /krow/, is usually written krówa because this is the way it sounds, though the terminal “a” is not phonemic and is often not expressed in sentence sequences.

See Appendix 4 for a more complete description of several of the linguistic aspects outlined above.


Sources for this chronology are Hemming (1987), Nimuendajú (1946), F. Ribeiro (1815, 1819a, 1819b), several Indian service agents (especially Olímpio Cruz) and Olímpio Fialho, historian at Barra do Corda, the missionary-linguists Jack and Josephine Popjes, and, most significantly, the Canela and Apanyekra research assistants (especially the older Mïïkhrô).



First contacts with Brazilian military detachments

ca. 1710

Bypassed by pioneer fronts moving south of homelands across Maranhão from the Parnaíba to the Tocantins


Defeated in battle by the Cakamekra


Pacification; accepted protection of Brazilian garrison at Pastos Bons


Fled protection at Caxias because of smallpox and returned to homelands

ca. 1816

Hid by a spring in an Alpercatas range valley (Vão da Serra), and were peacefully brought out of hiding by the military


Settled at the Porcos stream entrance to the Corda River and experienced miscegenation


Chief Luis Domingo Kawkhre assumed leadership, the first chief to be appointed by backland political authorities

ca. 1835–1845

Summoned by the local Brazilian authorities to fight in the Balaiada wars against the Brazilian Cabanagem Rebellion-related uprising in the backlands


Summoned by the local Brazilian authorities to fight against the Gamella Indians

ca. 1850

Zé Cadete Palkhre became chief

ca. 1870

Experienced relative affluence under Chiefs Coronel Tomasinho and Major Delfino Kô?kaypo in the Escalvado villages

ca. 1894–1903

Boys sent to study in the convent in Barra do Corda

ca. 1898–1901

Allowed the Cakamekra to join their tribe


Summoned by Barra do Corda authorities to fight against the Guajajara uprising at Alto Alegre


Graduated the age-set of Khà?po in an Escalvado village

ca. 1903

Tried and executed Francelino Kaawùy for a witchcraft murder


Split their tribe after the execution and relocated the two parts away from the Santo Estévão stream.


Experienced times of economic insufficiency while away from the Santo Estévão stream and its better gallery forests

ca. 1903–1922

Graduated the age-set of the older Mïïkhrô

ca. 1913

Intimidated by the ranchers’ massacre of the Kenkateye-Canela


Joined the two parts of the tribe, ending the schism


Suffered the great drought and consequent hunger


Returned to the Santo Estévão stream and relative economic self-sufficiency

ca. 1922

Graduated the age-set of the younger Ropkhà and ended the practice of age-set ceremonial marriage

ca. 1923

Received the six visits of Curt Nimuendajú


Graduated the age-set of the older Kaapêltùk and moved from Ponto to Baixão Prêto


Suffered smallpox, the death of the older Ropkhà (Faustino), the new schism in the tribe, and the move away from the Santo Estévão stream and from relative economic self-reliance


Mended their tribal schism by allowing Nimuendajú to rejoin the two parts of the tribe at the Raposa stream


Accepted the arrival of the first Indian service agent (Castello Branco) with his family to live next to the Canela village


Ceased to practice the uncle-nephew plaza hazing ceremony and the custom of childless women sleeping in the plaza with men


Moved back to Ponto and the Santo Estévão stream and returned to relative economic self-sufficiency


Benefitted from the protection of the great Indian service agent Olímpio Martins Cruz; last period of economic self sufficiency


Graduated the age-set of Chief Kaarà?khre


Learned from the Indian service school teacher Dona Nazaré, who taught six youths to write


Sent the younger Kaapêltùk and Hakhà, as students, to live with the Indian service in São Luis


Graduated the age-set of the younger Kaapêltùk


Lost the last strong chief, Doroteo Hàktookot, and began the era of Chief Kaarà?khre


Divided the tribe into the villages of old Ponto and Rodeador

ca. 1952–1954

Divided the tribe into the villages of Baixão Preto and the new Ponto


Were dismayed by the termination of the Indian service’s policy of providing extensive goods and thereby recognized the end of their mythical accul­turation contract set up by the culture hero Awkhêê


Received the first visit of ethnologist William Crocker


Graduated the age-set of Kôham


Transferred the power of the Pró-­khãmmã to the age-set of the older Kaapêltùk


Suffered from the messianic movement of Khêê-khwèy, the massive attack of backland ranchers, and the forced relocation to the dry forests of Sardinha where they did not adjust ecologically or psychologically


Men shamed by Guajajara women into wearing clothing all the time, and both sexes learned to make traditional artifacts for external sale


Experienced the anti-alcoholic “conversion” of Chief Kaarà?khre


Received training; three students of Dona Nazaré began writing diary manuscripts


Formed temporarily four competing cerrado villages in the Campestre, Escalvado, Ponto, and Baixão Prêto areas


Received the SIL missionary-linguist Jack Popjes, with his long-range community development program


Returned officially to their cerrado homelands


Rejoined the five tribal segments in the Sardinha, Campestre, Ponto, Baixão Prêto, and Escalvado areas in the present large Escalvado village


Construction of a road bridge at Ourives (halfway point) to enable army vehicles to move directly into the Canela region to protect the Canela position after their return to their homelands


Received the Indian service agent Sebastião Pereira


Construction of three “permanent” large Indian service buildings in Escalvado: post, school house, and infirmary

ca. 1970–1974

Began high population growth after near elimination of endemic infant and childhood dysentery


Rejoiced in the apparent demarcation of their lands (legal in 1978)


Completion of the direct road from Barra do Corda to their Escalvado village


Graduated the age-set of Kôyapàà


Installation of a gasoline generator for electricity at the post buildings, which supplied light and two-way radio transmission to summon aid

ca. 1973

Converted by Indian service agent to a belief in pharmacy medicine to cure and nearly eradicate tuberculosis

ca. 1974

Experienced first divorce in which children were involved that was granted by the service and the tribal council


Completion of a truck road from Escalvado to the village of Porquinhos of the Apanyekra


Served by an effective Indian service team: the agent Sebastião Pereira, the teacher Risalva Freire de Sá, and the nurse Luzanira Gieira de Araojo


Influenced by an Indian-self-conscious­ness and human-rights-oriented Indian service official in Barra do Corda


Final departure of ethnologist William Crocker and the termination of the diary manuscript and tape program [2001: Research begun again in 1999, continuing the diary program into the present.]


Transmitted the power of the Pró-khãmmã to the age-set of the younger Kaapêltùk

ca. 1981

Installation of an Indian service store for buying material artifacts for resale in outlets throughout Brazil, facilitated self-sufficiency

ca. 1981

Deposition of Chief Kaarà?khre by the new Pró-khãmmã age-set of the younger Kaapêltùk and his replacement by six new chiefs in succession due to political instability

ca. 1981–1987

Benefitted by an extensive farm project in the Pak-re area, under the leadership of the younger Kaapêliùk and financed by Vale do Rio Doce project funds


Split into five communities (total population about 800) with Escalvado hav­ing no leader and being almost abandoned, Pak-re having the largest number (~250) under the younger Kaapêltùk, Dois Riachos having an appreciable number under the de­posed Chief Kaarà?khre, Campestre having few in number under the current chief, the youngest Mïïkhrô, and Os Bois having few in number under a former chief, Krôôpey


Benefitted by the stability of having the same dedicated Indian service personnel at Escalvado since 1978, who provided truck roads averaging 15 km to reach 13 outlying farm communities (Map 3), enabling continuity of post services: leadership, protection, medicine, and schooling


Stabilized politically by the appointment to the chieftainship of the younger Kaapêltùk, who continues to hold this position in mid-1989

Population reached 903 (1 March 1989) according to official census and list of names made out and sent by Sebastião Pereira through Jack Popjes. [2001: Population about 1300.]



<<back  table of contents  next>>

[top of page]