Mortuary Practices and the Study of Prehistoric Social Systems Author(s): Joseph A.

Tainter Reviewed work(s): Source: Advances in Archaeological Method and Theory, Vol. 1 (1978), pp. 105-141 Published by: Springer Stable URL: http://www.jstor.org/stable/20170131 . Accessed: 11/12/2011 14:14
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Springer is collaborating with JSTOR to digitize, preserve and extend access to Advances in Archaeological Method and Theory.

http://www.jstor.org

Mortuary Practices and the Study of Prehistoric Social Systems

JOSEPH A. TAINTER

a variety a review might in which of ways be written. in a chronological is one possibility; of research framework Organization discussion of substantive results of research would be another. The pres ent review will follow a different course. In a field in which the literature a review or results has proliferated substantive detailing considerably, or as a listing of research results, might become ganized chronologically, There are tedious. More important, such approaches might not yield a clear picture of the strengths and weaknesses of the concepts and analytical methods research. This review is primarily oriented toward a discussion underlying are the extent to of these latter topics. Among the points to be discussed which different of the social significance of mortuary remains, concepts and the analytical methods that derive from these concepts, may be ex to augment our ability to study social variation and change. If the pected of mortuary does not ultimately contribute archaeological study practices to the general anthropological then our research study of social change, to be of limited interest. This review may ultimately is, then, not prove so much with the past accomplishments concerned of mortuary studies, as with the present and the future. With these points in mind, it is hoped that the present currently analysis, how mortuary review will show that the study of mortuary practices to increasingly and quantitative methods of aspires objective to concepts to the recognition of that are increasingly sensitive remains may reflect social phenomena, and to explicit

105
ADVANCES INARCHAEOLOGICALMETHOD AND THEORY, VOL. 1 Copyright ? 1978 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-003101-9

106

JOSEPH A. TAINTER

evaluation tuary

of concepts

and methods

by reference

to ethnographic

mor

systems.

MORTUARY BASIC CONCEPTS PRACTICES:

framework that underlies the use of mortuary data for conceptual inference has been set forth by Saxe (1970) and Binford (1971). uses elements of anthropological role theory developed approach a set of terms defining has developed (1965). Goodenough by Goodenough elements of social interaction that have archaeological implications. term social to what otherwise Goodenough's identity corresponds of social identities might be chief be called a social status. Examples might The social Saxe's or commoner, professor ter, and the like. When or student, mother or mother's brothers' daugh two or more in a proper social identities engage The parties to a social this is termed an identity relationship. relationship, do not usually interact in terms of only one social identity at a relationship time. An individual might manifest the identities of faculty advisor, friend, and coauthor, all in the course of a single interaction. Not all identities are appropriate for each interaction. The that a person could manifest as appropriate of several social identities selected for a given composite

interaction for the interaction. is termed the individual's social persona The kind of social persona that a person may manifest for an interaction will be determined characteristics of the social by the organizational a set of social personae will reflect, and contain informa system. Hence, of a particular the organizating tion about, (Saxe society principles In of this fact is useful for archaeological inference. 1970:7). Knowledge on differing levels of complexity, social identities will societies organized for them that it is possible vary as to the number of identity relationships to have. In egalitarian societies infants will have few social identities, on this principle, Saxe whereas adults will have acquired many. Drawing find infants buried in a manner (1970:8) points out that, if archaeologists by some adults, a indicating a social persona larger than that possessed indicated. of social ranking by birth is probably principle to With this array of concepts derived from role theory, Saxe proceeds to the analysis of mortuary data. We formulate a comprehensive approach to discuss Saxe's work at many points in this review. shall have occasion for concepts confirmation relating to mortuary practices Ethnographic is crucial for archaeology. Saxe (1970) and other practitioners (Tainter effort to considerable 1975b; Goldstein 1976; Vehik 1975) have devoted mor the most this area. But perhaps important survey of ethnographic In this case the word "impor is that of Binford (1971). tuary procedures

PREHISTORIC

MORTUARY

PRACTICES

107

a substantial body tant" is used not simply because Binford has consulted of ethnographic but also because the results of his survey literature, confirm beyond serious contention the argument (still rated skeptically by in terms in mortuary some) that variability practices must be understood not in terms of variability in the form and organization of social systems, of normative Binford modes set of behavior. out to test the following two propositions: (1) that "there should be a high degree of isomorphism between (a) the complexity of of the status structure in a socio-cultural system and (b) the complexity as regards differentiated treatment of persons ceremonialism mortuary status positions," should be a and (2) that "there different occupying characteris the nature of the dimensional between strong correspondence treatment and the ex tics serving as the basis for differential mortuary ar societies criteria employed for status differentiation among pected on a scale from simple to complex" this (1971:18-19). Rephrasing ranged last proposition, Binford argues that in an ethnographic survey we should hunters and gatherers find that among egalitarian age and sex should serve as bases for mortuary whereas distinction, among more commonly of agriculturalists social position societies (varying indepen complex should more frequently be the dently of age, sex, and subgroup affiliation) basis for distinctions treatment. in mortuary Binford acknowledges that his ethnographic sample is not ideally struc For example, since it was tured and suffers some operational problems. not possible to directly measure social complexity from the ethnographic Bin this was done indirectly by noting means of subsistence. literature, on the "generally ford bases this procedure correlation between accepted forms of subsistence and societal complexity" (1971:18). Sub production sistence mode was grouped into four categories: hunters and gatherers, shifting agriculturalists, this problem Despite considered the following
1. The differentiated the specific dimensions

settled agriculturalists, and pastoralists. the results obtained were meaningful. Binford to be satisfactorily demonstrated. points
recognition complexity in of

society 2. The number

mortuary as measured

of the social persona commonly given with the organizational vary significantly forms of subsistence practice. by different of dimensions of the social persona commonly given ritual complexity take, vary

with the organizational rituals varies mortuary significantly as measured forms of subsistence practice. by different ritual 3. The forms, which in mortuary differentiations the dimensions of the social persona symbolized

in recognition of the society, with

significantly

[1971:23].

Binford form and

concludes: structure

"These which

the generalization findings permit characterizes the mortuary practices

that the of any

108

JOSEPH A. TAINTER

society are conditioned by the form and complexity of the society characteristics itself" (1971:23).

of the organizational

PROBLEMS INTHE INTERPRETATION MORTUARY OF REMAINS

In a discussion which suggested that archaeologists should be cautious in the interpretation of mortuary remains, Ucko (1969:273) noted a case to as "an archaeologist's The Ashanti which he referred nightmare." follow a general rule that an interred body should not face the village. But turns itself after burial the body Ashanti say that immediately to face the village. this but not all, Ashanti cope with Some, that it will turn dilemma by placing the body facing the village, knowing itself around to the correct orientation. is an easy one to play with The game of "archaeologist's nightmare" Au from western data. Consider the following example ethnographic around stralia.
The the bones of adults into three have a unique custom of dividing Lyne River people in and ribs are placed bundles. The arms, collar-bone, shoulder-blades, shins, hands, one. The thighs, feet, hips, spine, and teeth are bestowed in a second. The knee-caps, to the third. One bundle is taken to the top of spine, and jaw are assigned as a spirit child by his or her father. Another was first "found" the deceased pool where third is The cord the deceased's is buried. bundle is interred at the place umbilical to the place where initiated or the women the man was taken by a mother's brother rubbed memorates she first with charcoal after birth. A man's skull his first killing of a kangaroo. crawled 1949]. [Davidson A woman's the is put under skull is carried stone which com where to the place

some

breast-bone,

cases. Perhaps to such ethnographic are many possible reactions unre the easiest would be to simply shake our heads, mutter something remains is impos and conclude of mortuary that interpretation peatable, not unknown in the archaeologi sible. Such a reaction is, unfortunately, There is to make sense out of The goal of any science, however, cal profession. to find the common factors diversity. chaos, linking apparent seeming in this chapter have ever is reviewed None of the persons whose work is simple, or implied, of mortuary remains that the analysis claimed, or easy. The diversity that have been of approaches straightforward, If the the contrary. indicate for mortuary clearly analysis developed factors conditioning ritual, and its archaeological mortuary preservation, are complex, this circumstance should not be thought of as a detriment, of mortuary that complexity To suggest but rather as an opportunity. that we is to suggest ritual renders burial data inscrutable ignore an

PREHISTORIC MORTUARY

PRACTICES

109

opportunity
ord.

to understand

a major

component

of the archaeological

rec

will clearly nightmares" "archaeologist's Any listing of ethnographic ritual is a complex that the form of a mortuary indicate interplay of ritual, of transformation factors. The archaeological social, and environmental at to attempts level of uncertainty remains adds yet another mortuary concen in this chapter The literature reviewed past societies. modeling trates on the social cism to be voiced of mortuary dimensions Perhaps one criti practices. is that the other factors that condition mortuary prac as have social dimensions. For tices ought to be treated as systematically who are primarily concerned with the nature of past those archaeologists of the ritual and treatment of systematic the advantage societies,

for aspects of mortuary practices would be the potential re that genuinely in archaeological data, those variables discriminating, within the state this can be accomplished flect social factors. To a degree, but not to a sufficient degree as to warrant of our present knowledge, environmental further ritual provides training are familiar with examples of people who bury their dead in trees, in rivers, at sea, and so forth. Although there are sound reasons (to be discussed shortly) for remains to reliably reflect the social information believing archaeological the archaeological communicated ritual, nevertheless through mortuary a sampling problem. At remains clearly presents of mortuary recovery from mortuary remains re societies tempts to characterize prehistoric in a about all kinds of burial practiced that we obtain information quire in many cases. Rather than find this This is obviously impossible society. There a discouraging limitation, we might see it as a genuine opportunity. remains are not in the extent to which mortuary well be patterns may to obtain posi itmay be possible recoverable Similarly, archaeologically. as absence of burials. For tive information from such negative conditions the extent to which certain age or sex classes are absent from an example, social factors. mortuary archaeological population may reflect significant as the that such a negative condition shall see later in this review We is a very informative characteristic. absence of formal cemeteries of ar In short, the complexity of mortuary ritual, and the problems rather than det should be viewed as opportunities chaeological analysis, of mortuary remains riments. Such a view does not imply that analysis a simple procedure. be considered There are good reasons to believe may ritual in of mortuary that common factors condition the social dimensions common factors are of a highly contexts. However, these varying cultural abstract nature; their archaeological application will vary with each case has been solved. that the problem complacency of mortuary The archaeological transformation with anthropological All archaeologists nightmares.

110

JOSEPH A. TAINTER

the concepts Thus, being considered. taken as a tool kit for archaeological for deriving interpretive principles

should not be but rather as a base interpretation, for each individual case. appropriate here

to be discussed

THE SIGNIFICANCE MORTUARY OF DATA

To evaluate the usefulness of mortuary two data for social modeling, criteria may be discussed. These are the range of social information that may be derived from mortuary remains, and the reliability of burial data as indicators of social phenomena. In his discussion (1970:6) between noted of the archaeological of role theory, Saxe application the occasion of death will involve an interaction the deceased person and many of the persons with whom he or that

she had engaged in identity relationships during life. It can be seen then that death and mortuary ritual call forth a fuller representation of an individual's various social identities than does any occasion during life. a the archaeological record of mortuary ritual should contain Hence, in a past about the social identities present greater range of information of information. Additionally, since society than does any other category individuals acquire social identities by virtue of membership in the struc tural parts or components of a social system, mortuary ritual will simulta information about the nature of a past society. neously convey Indeed, to the extent to which a mortuary contains individuals who held population membership lation can of a society, that mortuary popu components to reflect the structure of the extinct expected society of (Tainter 1977b:70). There does not appear to be any other category data for which this claim might ad be so confidently archaeological in the various be that mortuary Given ritual has such potential for extensive representa we must consider whether tion of social phenomena, mor archaeological mor tuary remains will reliably reflect the information conveyed through is the question raised by the tuary ritual. This, of course, continuously an answer to this ques skeptics and critics of social modeling. Deriving tion will involve a discussion in later of topics that will figure prominently of this review. of Saxe's work deals with the application to of formal analysis an approach pioneered in the middle Brown 1960s (but data, mortuary by not published until 1971). Formal analysis is a technique used to evaluate and display the combinations of mortuary found in a mortuary attributes as a branch domain. Such combinations of attributes are often displayed or key, as shown in Figure 4.1. ing diagram, Much sections

vanced.

PREHISTORIC

MORTUARY

PRACTICES

111

Head public Corpse displayed Head

in retained location

retained

in

private Formal /burial

dwelling

Left hand retained in private dwelling not Corpse displayed Right hand in private Deceased/ person near Disposal cultivated field Disposal on land Disposal \Casual disposal Disposal Disposal on water * Disposal Figure 4.1. Key diagram of a perfect tree. in sea in river in forest retained dwelling

tree. A perfect tree is an absolutely 4.1 is a perfect in that the decision made at any one contrast set (for the person merits formal or casual treatment) automati example, whether contrast sets. Thus, if the choices available in subsequent cally prescribes a person is to be disposed the corpse the question of whether of casually, is inapplicable. should be publicly displayed Instead, the next choice to be is whether should be on land on on water. Simi the disposal prescribed the fact that disposal in a forest redundantly larly, disposal emphasizes was on land, and was carried out casually. The key in Figure redundant structure to a perfect tree is a perfect paradigm. In a perfect contrast all attributes are independent. Choice of one attribute does not paradigm or limit subsequent contrast sets. Hence, is the redundancy prescribe zero. A perfect paradigm is illustrated in Figure 4.2. a key represents a tree or a for measuring whether Techniques are available fron the field of information In a situation paradigm theory. In polar

112

JOSEPH A. TAINTER

Grave Cremated No Articulated Grave Not cremated No

associations

grave

associations

associations

< Deceased, person

grave

associations

Grave ^Cremated No Disarticulated* Grave Not

associations

grave

associations

associations

cremated<Q ^No grave associations

Figure

4.2.

Key

diagram

of a perfect

paradigm.

of complete all attributes there combine with all others; nonredundancy, In information is complete randomness in organization. theory, this is a In a state of high redundancy, situation of maximum the combi entropy. and nations of attributes are highly determined; there is high organization a tree or a a key represents low entropy. whether So to determine its degree of entropy. paradigm we measure In a situation of maximum entropy all attributes will combine randomly If the number with all others. of possible combinations (termed sig will randomness is denoted as 5, then the situation of complete nificata) be expressed as S ? Ca x Co x where take. C
Coo

each attribute may values the number of different represents 4.1 and 4.2, there are two values for each attribute, with (In Figures the exception of the attribute Deceased person.) The value 5 will indicate the maximum number of possible burial modes and might be referred to as ? The amount of information in 5max can be measured E = log2 5max entropy possible in a key. as

Here

E will

represent

the maximum

PREHISTORIC

MORTUARY

PRACTICES

113

To measure entropy

actual

significata) actually of 5actual will

entropy we tabulate This will observed. be e =

the number be referred

of burial modes (or to as 5actUai- The

log2 S actual

of actual entropy. where e is a measure as Relative (RE) is measured entropy RE And conversely, to measure = elE apply the formula

(R) we redundancy R = 1 RE

for a perfect tree it will For a perfect paradigm R will equal zero, whereas have a value of 1 (Saxe 1970:102-107). are many of information There of Saxe's application implications to mortuary For the moment we shall consider only the remains. theory 1975b: 107-109). (Tainter following a communication in which certain ritual is basically system Mortuary are employed to convey about the status of the information symbols deceased. As in any communication the messages system, generated ritual are subject to noise, which may induce errors or through mortuary or inject extraneous Since noise can material into a message. distortion, or significance it is necessary for reliable of a message, alter the meaning communication

in to develop codes that allow transmission of information into the presence of noise. This is accomplished by building redundancy the code (Shannon 1949:75). a some symbols are used jointly to designate As we have seen, where is low. And number of disposal the redundancy types (as in a paradigm), to only a single burial type where each set of symbols pertains exclusively is high. In the former situation, archaeological (as in a tree), redundancy individuals will be difficult, because of socially distinctive identification In situations of certain symbols will pertain to a variety of burial modes. the opposite will be true. More high redundancy important for our present in mortuary where purposes, among the symbols employed redundancy ritual is high, the reliability of the archaeological record will be high. This is so because of mortuary in the ritual that are preserved the elements archaeological was conveyed that reaffirm the same message redundantly of the ritual. Hence, the ar elements by nonpreservable ritual will, in situations of loss of certain aspects of mortuary record will

chaeological not result in loss of information about the social charac high redundancy, teristics of the deceased individuals. the equations the amount of detailed above, Saxe investigated Using

114

JOSEPH A. TAINTER

Bontoc domains in three ethnographic (Ashanti, mortuary redundancy his sample was small, the re and Kapauku Although Papuans). Igorot, values Saxe (1970:230) found redundancy sults obtained were meaningful. and .88 (Bontoc .834 (Ashanti), of .764 (Kapauku Igorot). Papuans), These values are consistently high, indicating tree-like keys, and suggest ritual as a communication that mortuary system may universally ing employ a highly redundant code. This finding suggests a high degree of to the information in respect record for the archaeological reliability of the dead. communicated through disposal

STUDIES THE USE OF SOCIALTYPOLOGIESINMORTUARY

in the study of prehistoric societies has been One of the basic problems societies the development of scales on which may be archaeological used are The scales most frequently for comparative purposes. placed from the work of Service derived from ethnology, (1962) and primarily has since retracted his typology Fried (1971:156-157), (1967). [Service to employ continue it.] These scales aspire although many archaeologists is developed in that a societal typology to an ordinal level of measurement structural to increasing degrees of are ranked according of societies for organiz and increasing numbers of mechanisms complexity ing populations. as analogues for archaeological The use of evolutionary typologies studies. Some examples will suffice to has dominated mortuary societies derived when the types of conclusions illustrate evolutionary usually are employed. typologies in which kinds
. . . data from the cemetery at Rinc?n suggest ... an egalitarian society [Stickel

1968:227].
. . .the Mrn-27 was an in which rank seems most ascribed cemetery likely to reflect a society structure element of the social 1970:22]. [King important . . . existed . . . social stratification at least as early as Chumash the Northern among the CCo-308 for Middle Horizon components the Late societies [while egalitarian sense in Fried's [Fredrickson society is consistent Horizon wit

1971:16]. [Tainter of mortuary patterning the patterning with predicted The nesses] the emergence [The mortuary

500 a.D.

of a rank

at Spiro] fulfill some of the requirements patterns a chiefdomship called 1971:102]. [Brown organization to be It is suggested that a chiefdom should continue model culture the Dallas 1974:251]. [Hatch archaeological involving have [The Moundville as their Phase premises data fulfill] mortuary that the Moundville Phase

1974:62,64]. level of of an adaptive in research which [Peebles

employed

of hypotheses the expectations a ranked was society

1974:191].

PREHISTORIC

MORTUARY

PRACTICES

115

listing could continue much further, but the point has certainly of made that the archaeological study of the social dimensions on evolutionary has been overwhelmingly mortuary dependent practices some discus to devote This being the case, it is appropriate typologies. sion to the logic and the limitations of this approach. This been have The ethnologists who have developed evolutionary typologies and have utilized social variables as dichotomous, largely conceptualized as the basis for abstracting societal "types." such dichotomies Typical of are statements such as the following: evolutionary typologies
Band society, out of which tribes grew, was for egalitarian special [Service purposes is personal?charismatic?and only 1962:114]. in tribal society

Leadership [Service Tribes bands, Apart A

are not, like but chiefdoms sodalities, though, by pan-tribal for limited purposes sodalities [Service 1962:165]. they may have a few minor division of labor [in rank societies] from age and sex, there is little significant society equal is one access in which members of the same that sex and equivalent life [Fried age status

1962:114]. are integrated

[Fried 1967:129].
stratified do not have to the basic resources sustain 1967:186].

It is combinations that are used to define attributes of such dichotomous are conceptualized that categories societal typologies. of societies Given as combinations it would follow that the means by of discrete attributes, or archaeologi which a newly discovered (whether ethnographic society can be assigned to a proper evolutionary slot is by a process of keying cal) the society, on its various dichotomous until the final combina attributes, tion of elements its appropriate indicates designation. can indeed aside the question of whether social characteristics Leaving we shall consider be considered the archaeological discrete attributes, are to be assigned If prehistoric of these principles. societies application to an evolutionary slot by means of an identifying then the ar key, chaeologist identifying to the analysis of data with a list of criteria for should proceed to the keying each dichotomous attribute that contributes Such a process is logically intrinsic to evolutionism, but is not an evolu within Instead, applied. archaeologists working

procedure. consistently tionary framework regularly abridge this keying procedure by identifying a limited number of social characteristics, most often rank differ only for ences, and from these inferring the appropriate typological designation the society in question. The implicit assumption in this approach must be that the dichotomous attributes societal types are so strongly defining and so highly redundant, that identification of one implies all associated, as well as the abstracted the others, societal form they collectively desig nate. This assumption to rigorous testing, and so has never been subjected must be viewed with caution.

116

JOSEPH A. TAINTER

One recent archaeological study that has avoided the pitfall of determin a typological for a prehistoric designation society on the basis of only ing one or two social variables of social or is Renfrew's (1973) evaluation Renfrew that the large earth in neolithic Wessex. proposes ganization works that typify the later portions of this time period test this hypothesis, Renfrew To of chiefdoms. reflect the existence abstracts from the formulations of Service (1968) (1962) and Sahlins He lists these as follows: of chiefdoms. twenty defining characteristics and monuments
(1) a ranked society the redistribution

(2)

of produce

organized of the

by

the chief

(3) greater density population in the total numbers (4) increase (5) increase in the size of

individual

society residence

groups or borders statuses activity

(6) greater productivity territorial defined (7) more clearly (8) a more integrated society with (9) centres (10) (12) (13) frequent relation which coordinate ceremonies

boundaries

a greater number of sociocentric as well as economic social and religious serving situation to some wide social purposes

and rituals

(11) rise of priesthood

to a total environmental redistribution)?i.e. not only regional in large favouring specialization diversity also through in production (and hence ecological or ecological but

of the pooling endeavors. cooperative for agricultural work of public and deployment labor, sometimes (14) organization or pyramids for building and/or (e.g. temples, irrigation) temple mounds, specialization, individual skills (15) (16) potential (17) reduction in craft improvement for territorial of internal specialization expansion?associated strife with the 'rise and fall' of chiefdoms with

or groups in the society associated of persons (18) pervasive inequality nent in fields other than the economic effective leadership, (19) (20) distinctive dress or ornament to back for those of high by status force [Renfrew no true government up decisions legalized

perma

1973:543].

with these indicators, Renfrew proceeds with an analysis of the that the neolithic Wessex data, eventually concluding archaeological of these characteristics or inferred presence of the majority documented of chiefdoms. the existence indicates in many ways, is satisfying Renfrew's including not only its analysis that the chief stated recognition but the explicitly relative completeness is a catchall that lacks specificity. dom concept Indeed, Renfrew suggests "will have to make way for, or be refined to that the chiefdom analogue is (1973:557). This realization yield, subtler and less inclusive concepts" remain. of Renfrew's but some disturbing analysis aspects refreshing, Al characteristics. is his handling of the list of chiefdom Most serious Armed though many linkages can be discerned among these twenty attributes,

PREHISTORIC

MORTUARY

PRACTICES

117

that Renfrew the reader is still left with the impression regards a chiefdom seems as a list of traits. Renfrew to but not as an adaptive system, advocate what might be termed the checklist approach to social modeling. to has accepted his own challenge In a subsequent work, Renfrew con and less inclusive for the chiefdom "subtler substitute analogue into two of chiefdom the category is to subdivide cepts." His solution varieties: differential little evidence of with (chiefdoms chiefdoms group-oriented access to wealth, of communal evidence but considerable that display material (chiefdoms chiefdoms labor), and individualizing of monumental indications of ranking, but only a minimum construction) is careful to point out that this is not intended (Renfrew 1974:74). Renfrew

the to be an exercise Yet it is not hard to envision what in typology. this to its logical conclusion, outcome of this approach might be. Taken will bring us an solution to deriving "subtler and less inclusive concepts" as in the evolutionary of new pigeonholes endless proliferation typologies, chiefdom was not that his or her prehistoric each archaeologist discovers filled the literature may become quite like all the others. And ultimately this or that set of archaeological with endless debates concerning whether 32b! Clearly the data reflects Chiefdom 32a or Chiefdom Type Type will lead nowhere of "subtler and less inclusive concepts" development as long as such concepts of new labels to apply focus on the proliferation over our research effort worrying to prehistoric societies. To concentrate what to call a past society is a waste of that effort. If one of our objectives on the is to study social variation and change, then we should concentrate social characteris of truly quantitative scales for measuring development and will be dis have been developed tics. Such quantitative methods cussed shortly.

DATA OF CLASSIFICATION MORTUARY

data is the problem of of mortuary basic to any analysis Certainly The purpose of classifying mortuary data is to isolate clus classification. ters of burials, which can be interpreted as socially distinctive. The use of has been dis for this purpose Saxe formal analysis 1971; (Brown 1970) Saxe has taken the techniques of formal analysis and cussed previously. a system of interrelated the social di concerning developed hypotheses mensions
ing. The of a Given Components Disposal the Resultant Combinations Domain Representing Cooperate Different in a Partitioning Social Personae of the

of mortuary

practices.

Among

these hypotheses

are the follow

Universe,

[Saxe

1970:65].

118

JOSEPH A. TAINTER

Putting it in other words, Saxe a different social persona.

Within Fewer [Saxe a Given Positive 1970:69]. Domain Components Personae

postulates
of Lesser Significata

that each burial

type represents

Social

in Their

to Manifest Tend Significance to Others, and Conversely Relative

The

of is that social personae last hypothesis in identity numbers of persons engage larger higher significance of lesser sig than will personae (usually status obligations) relationships in mortuary will be symbolized These nificance. identity relationships reasoning behind this will ritual,

and reflected as positive (or in other words, components positive of higher significance in the burial domain. Hence, personae attributes) itmight be added will display more positive components. Parenthetically, for the former Saxe found that in his ethnographic survey support whereas hypothesis, These hypotheses will be useful archaeological the latter tested underscore clearly to evaluate how materials. ambiguously. the potential of formal analysis. been applied It to

the technique

has actually

Brown's of the use of formal analysis (original Following exposition was applied to several ar the technique 1966, published paper 1971), data sets in California 1969; Finnerty 1968; Decker (Stickel chaeological was used in one case involving the lower Snake et al. 1970; T. King 1970) and from River region in southeastern 1973). Viewed (Rodeffer Washington the perspective of Saxe's work, several of these studies display a common data, formal keys tend to focus problem. When applied to archaeological on variables variations that reflect idiosyncratic the classification process to individual burials. The resultant often yield classifications peculiar one burial individual formal each. When types represented by only to gain it is difficult isolate individual burials, classification procedures of and the organizing the structure information principles concerning can be that existed in past societies social systems. Social positions similar identified archaeologically by isolating sets of burials manifesting not by keying out individual burials. social personae, These results do not imply that formal analysis may not be negative or that Saxe's used archaeologically, may not be opera hypotheses should tionalized. The mortuary attributes used in a formal classification must be that such for selection be selected The criterion carefully. attributes variance. display common to formal analysis, mortuary As an alternative data might be classified are Such techniques the use of multivariate statistical techniques. through that variables sets of interrelated suited for the task of isolating ideally kinds of mul common when different variance. Unfortunately, display are applied to the same set of mortuary tivariate procedures data, the

PREHISTORIC MORTUARY

PRACTICES

119

results obtained may vary widely. This review is perhaps not the proper so we shall restrict the nuances of multivariate statistics, place to discuss to citing the results of an experiment in classifying ourselves mortuary data using a variety of statistical procedures (Tainter 1975a). The results of this experiment classification seemed to indicate that polythetic proce data because dures (Clarke 1968:37) are not entirely suitable for mortuary are not satisfactorily the clusters these procedures derived through tech In a set of experimental classifications polythetic homogeneous. of several rank members of burials containing as are available, set of techniques known grades. sub monothetic-divisive These procedures. procedures progressively at each in such a way that the resulting divide a population subgroups, are maximally A stopping rule is applied hierarchical level, homogeneous. to halt the subdivisive at the proper point. Of the various al process niques clusters yielded An alternative gorithms on which the hierarchical of attributes for deriving arrangement two of the most common are various is divided, the population statistic (Pee functions of chi-square 1972) and the information (Whallon data the latter bles 1972). In a test using Middle Woodland mortuary statistic gave superior results (Tainter 1975a), and it has proved useful in data sets (Peebles 1972, 1974; analysis of a large variety of archaeological available

Tainter 1975b;T. King 1976;Goldstein 1976).

STUDIES SELECTION DATAFORMORTUARY OF

data for use in mortuary The selection of categories of archaeological studies must be based on the variety of means by which social positions are symbolized treatment. Many archaeological studies have in mortuary in of symbolic forms which may be employed the diversity neglected infor and have assumed instead that the most significant ritual, mortuary As a mation may be derived from one data class: grave associations. framework has been built around the result, a rather elaborate interpretive Initiated by Binford (1962), this framework study of grave associations. to develop has been elaborated Stickel's Stickel (1968). by goal was a set of mortuary whether criteria for determining data reflect a society or ranked. that was For an egalitarian Stickel society, egalitarian that grave associations should reflect the following (1968:217) proposed characteristics: 1. A predominance cal sphere 2. Status symbols of artifacts that served by primarily individual in the technologi achievement

of the sort attainable

120

JOSEPH A. TAINTER

such status groups as exist, grave associations that distribute rather than groups individuals among differentially 4. Noninheritance of status symbols, potentially signified by inclusion of status symbols in graves 3. Within the members among should associations ranking, grave kel 1968:217), as follows: In contrast, of a society characterized by class in a contrasting fashion (Stic pattern

1. Increased of status symbols frequency 2. Possession of specific status symbols by groups 3. Considerable individual variation in mortuary associations 4. Inheritance of status symbols these concepts to the 1700 b.c. southern California Rinc?n Applying re Stickel that the mortuary concluded remains cemetery, (1968:227) flected an egalitarian society. to be a stimulus for research Stickel's into the social paper proved of California studies docu organization hunter-gatherers. Subsequent the evolution of complex, mented areas of California about beginning in several systems 1970, 1974, 1976; (T. King Tainter social systems, unusual 1971; Fredrickson 1974). Such complex are clearly documented in the ethnographic for hunter-gatherers, litera record. Late prehistoric ture, as well as in the archaeological cemeteries, seem to reflect the ethnographic for example, situation of differentially et al. 1970). ranked kinship groups (King 1969; Finnerty the florescence of productive research stimulated Despite by Stickel's some nagging questions about this approach remain. Much of paper, framework boils down to a distinction in Stickel's differentiation between of grave associations and differentiation in quality. In an egalitar quantity status positions ian society there will be few structural that need to be in items. Thus, grave associations symbolized through exotic material a society will be restricted, as Stickel proposes, to differentiations in such of burials in terms of quality or type of quantity of items. Differentiation status markers exist that will be minimal, since few status positions In a ranked society there are more structural require use of such markers. material status positions, and there exists a corresponding need for more kinds of In addition to differentiation items to denote status distinctions. in quantity of items, deceased members of a ranked society should also be in terms of the kinds of status markers the grave. distinguished placed-in ranked 2000 b.p. social

Of course, both egalitarian and ranked societies will be characterized by in both quantity and quality of material associations. burial differentiation But it is to be expected differ that in egalitarian mortuary populations

PREHISTORIC

MORTUARY

PRACTICES

121

in quality or type of grave association will be less pronounced entiation than in a ranked population. data that create doubts about archaeological There are ethnographic use of the concepts above. The northern California Modoc presented to the corpse, of their dead. In addition cremation the clothing practiced were placed on the pyre to be and personal possessions of the deceased consumed. items lost in this way. Any though, were valuable Rarely, individual attending the ceremony might remove an object from the pyre it with a quantity of shell beads, or perhaps render some replace For the archaeological service such as fire-tending 1963:116-117). (Ray In the case of the of such actions are significant. record the consequences the material Modoc items accompanying the deceased person are initially and differentiated and quality In Stickel's of materials. this would indicate a ranked here) seemingly not make its way of ranking would apparently into the archaeological in type and record. Instead, the initial distinction to mere differentia of grave associations is ultimately quality collapsed tion in quantity of shell beads, an archaeological that might characteristic type framework (as interpreted society. Yet such evidence as reflecting egalitarianism! studies have paralleled Stickel's Many archaeological approach by rely as the sole or primary class of data for deriving ing on grave associations social inferences 1968; (in addition to the studies cited above, see Winters interpreted 1971; Peebles 1971, 1974; Hatch 1974; Rathje 1970, 1973; Autry in terms of

be

Larson Shennan

1974; Baker 1974; Clark 1969;Randsborg 1974, 1975;Rothschild 1973;

enter the analysis 1975). In most cases in which other variables location of grave, etc.), these characteristics (type of interment facility, are not used as independent sources of information, or used in conjunc to derive social information. tion with grave associations Instead, grave are consistently associations information treated as the major source, with other categories of mortuary data treated in a wholly subsidiary fashion. There is no logical or empirical basis for this approach. Mortuary ritual is a process The nature of a symbol is such that the of symbolizing. between the form of the symbol and its referent is arbitrary, relationship Given or, at most, expedient. this, there is no intrinsic reason why social must be symbolized distinctions in a associations. Indeed, by mortuary recent survey of ethnographic 1974: 125), it mortuary systems (Tainter was found that the use of material to signify status distinctions inclusions was a decidedly minor practice, used in less than 5% of all cases (n= 93). This result alone suggests that the extensive reliance archaeologists place on grave associations is likely to be inappropriate. We are left then with the question, On which dimensions of mortuary

122

JOSEPH A. TAINTER

concentrate ritual should archaeologists To answer their analysis? this entail listing all dimensions of mor directly would essentially question or either archaeologically tuary ritual that have ever been observed, a Such a list would not only be ponderous, but, without ethnographically. valid interpretive it would be mean framework for each such dimension, on a the next section of this review will concentrate Instead, ingless. that seem most suitable for comparative discussion of mortuary variables
analysis.

MORTUARY STUDIES:CROSS-CULTURAL PERSPECTIVES

If the study of past social systems is ever to aspire to the study of social must develop then archaeologists the ability to model prehistoric process, in in ways that yield comparable results. Two of the factors societies are explicit con in the development volved of comparable social models sideration of the dimensions of a social system that are being investigated, the development of cross-culturally these dimensions. measuring Any social system is differentiated and valid criteria for isolating and

along a number of dimensions. this crucial dealing with past societies generally neglect Archaeologists that are most easily consideration, deriving only those social inferences obtained from a set of data, rather than specifying in advance which social a characterization dimension is the objective of the study. Blau provides of the concept
A dimension are

of dimensions
of differentiation

of differentiation:

organization .... subunits subunit structural

on the basis of which is any criterion the members of an ... ... or into ranks, or into divided into positions, formally ... or a A structural is either a distinct status official component . . . The term differentiation in the organization. to the number refers of components emphasis that are formally in original]. distinguished in terms of any one criterion

[1970:203-204;

the important concept of dimensions of differentiation Failure to consider to derive an impressive has led archaeologists about array of information that is of information but has rarely led to the production past societies, in one region of the Thus, where comparable. archaeologists working units (Longacre infer the existence of uxorilocal residence Southwest a society with discovers 1970; Hill 1970), an archaeologist working nearby to study so hereditary ranking (Clark 1969:82-83). Investigators wishing are left at a loss. and change in these adjacent cial variation regions that we should be explicit about the dimensions of a social Given we are attempting we must to model, tackle the more difficult system social information be derived from mor question, How may comparable

PREHISTORIC

MORTUARY

PRACTICES

123

social differentiation among the dead is tuary data? The means by which even among groups at burial varies considerably, living in symbolized the ar (Kroeber 1927). The problem confronting geographical proximity has, for this reason, been the lack of ability to develop general chaeologist remains that could be applied cross for interpreting mortuary principles culturally. This lack nearly every set of strategy for social attempt to produce a situation where of generalizing power has produced with a unique data must be approached mortuary inference. And yet we have seen that at least one an objective, valid strategy for mor cross-culturally We may tuary analysis (Stickel 1968) contained unforeseen shortcomings. or conceptually to search for analytical realistic ask, then, if it is feasible intended to produce frameworks information. comparable two social dimensions At present, have been singled out for the de These are (a) descent frameworks. of comparable velopment analytical Treating the first charac

and (b) rank differentiation. group differentiation the following teristic, Saxe proposed hypothesis:
To stricted Areas the Degree that Corporate are Attained Resources for

the Dead

(i.e., Lineal Ties the Exclusive Disposal

to Use and/or Control but Re Crucial Group Rights from and/or Legitimized of Lineal Descent by Means to Ancestors), Such Groups Will Maintain Formal Disposal of Their Dead, and Conversely [1970:119].

for it helps account for a number of is provocative, hypothesis that have long puzzled archaeologists. The factors determining questions or absent whether in a particular burials are present or, if location, or are clustered, present, whether randomly distributed, they dispersed, with certain types of features, such as houses, are undoubtedly associated Saxe's hypothesis that much of the variation in burial multiple. suggests for by social factors, and that we may use distribution may be accounted Saxe's More to gain certain types of social information. variations to derive com offers an opportunity Saxe's hypothesis specifically, de the number and nature of corporate information parable concerning cent groups making communities. up prehistoric Saxe initially tested this hypothesis against his sample of three ethno confirmation. More recently, Gold graphic societies and found consistent on a more extensive stein (1976) has tested this hypothesis sample of thirty cases. The results of this test seem to indicate (a) that the presence of formal disposal areas is consistently with corporate groups associated lineal descent and (b) that most, practicing though not all, such groups use areas. Thus, formal disposal areas will the presence of formal disposal has isolated individual corporate indicate that the archaeologist strongly will suggest, with a proba groups, and the absence of formal cemeteries bility that is high but less than unity, sort. This last interpretive ambiguity the absence of social groups must be kept in mind. of this such distributional

124

JOSEPH A. TAINTER

Two recent ness. Goldstein

of Saxe's hypothesis have shown its useful applications has found the concept in elucidating useful the (1976) of the spatial organization of two Mississippian in cemeteries significance the Illinois Valley. In the Kaloko located in the Hawaiian Is cemetery, of the cemetery lands, the spatial organization appears to be hierarchically as a arranged (Tainter 1976). The entire Kaloko community was organized

in the fact that the region contains a single single corporate unit, reflected settlements from adjacent temple, and is physically separated by unoc cupied zones and boundary cairns (Tainter and Cordy 1977). Ethnographi such settlements held the land on which cally, corporately they were situated. Hence, it is not surprising that Kaloko but a single, contains formal cemetery. At the same time though, both the cemetery and the into smaller units. There appear to have been community were subdivided a total of four residential a units at Kaloko, each of which maintained use. This subdivision area seemingly for its exclusive separate agricultural of the is reflected in a into smaller corporate groups larger community area there in the cemetery. Within similar subdivision the larger disposal are a number of subareas and bounded. that are discrete Each of these to one of the residential groups of subareas in the cemetery may correspond

two or more cemetery the community, or, alternatively, groups may have been linked to form a single residential unit. This second alternative is indicated by the fact that, although there are only four residential units, the seems to contain at least nine corporate descent groups (Tainter cemetery 1976). with some hindsight allows us to see the Saxe's hypothesis Coupling in earlier mortuary observed of distributional patterns significance studies. Chumash cemeteries coast, be along the southern California tween about 2000 b.p. and the historic period, seem to contain evidence of at least two differentially ranked descent groups: a higher ranking one in and a lower ranking group in the eastern part of the cemetery, from the perspective of 1969; Tainter (L. King 1971). Viewed segment and with archaeological Saxe's hypothesis, and ethnographic knowledge about social ranking in native California, these patterns might have been locational consid anticipated. More recently, T. King (1976) emphasized to demonstrate com erations the existence of politically differentiated the western munities of hunters and gatherers about 2000 b.p. in the southern Sierra foothills of Califronia. and from objective The analysis of social ranking similarly benefits discussed valid criteria for analysis. We have previously cross-culturally or attributes, or the components, the use of formal analysis for classifying From the use of formal analysis Saxe has developed domains. mortuary the following hypothesis

PREHISTORIC

MORTUARY

PRACTICES

125

Within Fewer

a Given Positive

Domain Components

Personae in Their

of Lesser Significata

Social

to Manifest Tend Significance to Others, Relative and Conversely

[1970:69]. The was discussed in the section of this reasoning behind this hypothesis of mortuary data. Although the the classification chapter dealing with is consistent, the hypothesis logic behind ethnographic testing proved In operational dubious terms, we have already seen (Saxe 1970:226-227). use of formal analysis that the archaeological tends to focus the classifica tion on idiosyncratic that isolate individual attributes, producing keys burials. An alternative to isolating rank distinctions is derived from approach Binford's that the form of a mortuary ritual will (1971:17, 21) observation be determined other factors, the size and composition of the by, among status responsibilities social aggregate to the de recognizing obligatory ceased. Binford proposes that such a larger array of duty-status relation of persons of high rank) will ships (which is characteristic to a larger amount deceased in of corporate involvement and to a larger degree of disruption of normal interment, on this, Tainter activities for the mortuary ritual. Expanding that both the amount of corporate and involvement suggested of activity disruption will positively to the amount correspond entitle the the act of community (1973) has the degree

of energy act. Directionally, in the mortuary (or labor) expended higher social rank of a deceased to greater amounts of corporate individual will correspond and activity disruption, involvement and this should result in the expendi ture of greater amounts of energy in the interment ritual. Energy expendi ture should in turn be reflected in such features of burial as size and

of the interment facility, method of handling and disposal of elaborateness the corpse, and the nature of grave associations. this reasoning, Reversing when sets of mortuary data cluster into distinctive levels of energy expen will signify distinctive levels of social involvement diture, this occurrence in the mortuary indicate distinctive act, and will reflexively grades or levels of ranking. The use of energy expenditure vide a solution to some dilemmas for determining rank grading may pro encountered Energy by archaeologists. of rank patterns in situations where grave

allows the analysis expenditure are not present. associations Indeed, was first formulated for the analysis dimensions of variation were

facility (Tainter 1973). Energy tive to interpretive that focus on only one dimension frameworks of such as grave associations. we In the case of the Modoc, mortuary ritual, observed that an initial contrast in the type and quality of items associated

the energy expenditure argument of a cemetery where the primary the size and elaborateness of the interment a preferable alterna expenditure provides

126

JOSEPH A. TAINTER

with

individuals was transformed into a situation where the archaeological contrasts would reflect only quantitative in grave associations. Such a situation confounds the interpretive framework specified by Stic as the analytical kel (1968). The use of energy expenditure criterion would record solve archaeological for rank differences this dilemma, record of Modoc mortuary would practices if the still be apparent were viewed from this

perspective. in mortuary The proposition ritual to the linking energy expenditure sam rank of the deceased has been tested on an extensive ethnographic In a set of 103 ethnographic cases, the energy expenditure argument ple. was once (Tainter 1975b). A selection of examples not contradicted from cases will illustrate the kinds of energy expenditure these ethnographic contrasts that are consistently found in the literature.

Complexity of Body Treatment

an example to illustrate variations in the complexity of body so many cases might be cited. but only because is difficult, "the the Salinan Indians, near San Antonio Mission, California, Among most distinguished dead were cremated, while persons of no particular Selecting treatment buried" (Mason 1912:167). A Papuan group in importance were merely a In contrast, in a sewn mat. the Fly River region buried commoners on a chief, a chiefs person would be exposed only child, or a respected and placed in a specially wooden the skull removed, decorated, platform, and the other and buried erected bones collected (Riley building, of Borneo practice a type of communities 1925:166, 170). Ngadju-Dayak burial associated secondary more Only the economically ritual (Miles 1965). Construction Two and Placement with considerable economic expenditures. are able to afford this successful households

of

the

Interment

Facility

contrasts in illustrate energy interesting examples expenditure the Padju Epat Ma'anyan these variables. Among the final rest Dayaks, an individual is in the tambak, ing place for the bones of a deceased carved box used by the bilateral descendants of the person who ornately founded it. The establishment of a new tambak is an expensive affair, but social and of an almost unique an individual such a achieves to disassociate social and economic himself from his level that he wishes the construction of a new tambak to be used natal group, he commissions himself and by his decendants District (Hudson 1966). In the Morobe by the body of a deceased individual was ordinarily placed in of New Guinea, to decompose After the ground and allowed for a period of five months. such an act symbolizes the attainment status by the founder. When economic

PREHISTORIC

MORTUARY

PRACTICES

127

that time, the interred. The the cliff varied left at the foot

skeletal amount

remains were taken to a cliff face where they were in disposing of the remains in of energy expended with the status of the deceased. Children and women were of the cliff, or placed on lower ledges. Important men were ledges, and the remains of the highest ranking members were carried to the highest ledges of the cliff (McWil

placed on higher of the community liam 1936). Extent

and Duration

of Ritual

Mortuary

Behavior

accounts of early Hawaiian mortuary practices, Two missionaries' both in rank dating to the same time period, clearly illustrate how differences in the extent and duration of ritual mortuary may be reflected behavior, and correspondingly in energy The two accounts relate expenditure. for community and for persons of chiefly outcasts, mortuary procedures two social classes rank. These the extremes of the Hawaiian represent In the early 1820s, Charles noted that social rank system. S. Stewart were often simply outcasts those with physical abnormalities) (including were removed from the living area of left where they fell, or, if necessary, one example the community. Stewart of such a means of described disposal
An

(quoted

in Bowen

1961:74).

within

a poor wretch came thus perished to our knowledge in which recently in the open lain uncovered for days and nights after having sight of our dwelling, in vain to his family, the hearing of his voice, still within air, most of the time, pleading he was dead, his body, instead of being buried, was for a drink of water! And when instance as to prevent the offense into the bushes, left a prey to the dogs who prowl through that would the district have arisen in the night!

so far drawn merely and from the corpse,

Ellis death

a starkly contrasting described of a local chief on the Island

level of social of Hawaii.

involvement

on

the

. . .the whole seldom their was

witnessed clothes,

a scene of confusion, exhibited and cruelty, neighborhood wickedness, in the most ran to and fro without barbarous The people society. and acting more than human beings; like demons every vice appearing

and almost every species of crime was perpetrated. were burnt, Houses even murder sometimes and the gratification of every property plundered, committed, base and savage restraint [1969:177]. feeling sought without practised,

that have been re of energy expenditure categories are material to the ritual, and contributions ethnographically human sacrifice. Ethnographic of these will not be detailed at examples this point, since the Modoc might be cited as an example of the former and the latter is relatively category, unambiguous. The ethnographic literature seems to indicate clearly, then, that energy Two corded additional expenditure in mortuary ritual is directly related to rank grading. As is

128

JOSEPH A. TAINTER

the archaeological of this idea is not however, application is measuring without One major problem energy expenditure, ambiguity. a meaningful in types and determining what constitutes energy contrast can only be of energy expenditure of burial. At present, measurement often the case, on an ordinal scale with burials ranked by greater or less In some cases decisions ordi (apparent) energy expenditure. concerning more to arrive at. A second, nal ranking are difficult and potentially is the extent to which archaeologically apparent dis important, problem accomplished rank differences. reflect The always numerous of mortuary prac examples as energy expenditure tices that might appear archaeologically variations, but in fact are not. The ossuary burials of the Huron are a case in point. individuals were periodically the bones of deceased the Huron, Among in energy expenditure literature provides ethnographic tinctions and placed gathered had the flesh removed who died shortly interments ossuary cemetery. Most from the bones, but this was not done to individuals before the reburial ceremony 1920:74-75). (Bushneil a Huron then display both articulated ossuary would a contrast that might be mistakenly inter individuals, in a communal

Archaeologically, and disarticulated in energy expenditure. preted as a variation a caution suggested Such an ethnographic previ example underscores concern to develop abstract generalizations it is possible ously. Although of such generalizations ing mortuary practices, archaeological application careful will always study of individual data sets. With require careful con to isolate many of the archaeological it should be possible scrutiny trasts

in energy expenditure that do not reflect rank differences. Ubelaker has properly noted that degree of skeletal articula (1974), for example, reflects the amount of time that tion in a Maryland ossuary probably death and ossuary interment (as among the Huron). With passed between offer a does the energy careful perspective application expenditure information about valid criterion for deriving comparable cross-culturally in varying contexts. Its usefullness is ultimately rank differentiation in which it has studies of archaeological trated by the number 1973; Tainter 1972; Brown 1973, 1975a,b, (Buikstra employed 1977a,b; Tainter and Cordy 1977; T. King 1976; Peebles 1974). illus been 1976,

MODELING QUANTITATIVE
and our goal in studying past societies must be to monitor Ultimately, a that we possess variation and change. Such a goal presupposes explain In archaeological social variables may be measured. scale against which of evolu studies the scales used for this purpose have largely consisted

PREHISTORIC MORTUARY PRACTICES

129

we found reason to doubt whether such Previously, typologies. are really suitable for this purpose. there are Fortunately, typologies to evolutionary As in and sensitive alternatives typologies. sophisticated to some earlier sections of this review, we find two basic approaches one based on formal analysis, the other linked to modeling: quantitative tionary the evaluation In developing Saxe practices,
The More the Less More

of energy expenditure. his comprehensive approach proposed

and More

to the analysis

of mortuary

the following
Evidenced

hypotheses:
in the Key Structure of the Domain, the Social the Conversely, Organization. the Social and the Less Egalitarian

Paradigmatic

the Attributes

Complex Tree-Like the Attributes, [1970:75].

Egalitarian the More

Complex

Organization The

for There To Be the Greater Will Be the Tendency System Simpler a Sociocultural a Linear in Significata, Number of Number Between of Components Relationship and the Social Contrast Sets Necessary To Define of the Sig Them, Significance and Conversely nificata; [1970:112].

The former hypothesis is based on the application of information theory as well as to key structures. We have already discussed this application, the procedure for measuring the extent to which any given key represents a tree or a paradigm. It is timely at this point to recall one of Saxe's that persons of higher social significance manifest previous hypotheses: more positive in their disposal domains components (positive attributes) than do less significant persons. If this is so, then not all attributes would as would be the case in a perfect to all burial modes be equally applicable persons of greater social (see Figures 4.1 and 4.2). Furthermore, paradigm treatment should evidence more and more kinds of mortuary significance as social com that are not applicable to less significant persons. Hence, the mortuary domain should become less paradigmatic. plexity increases, measure This hypothesis for social complexity. indicates a quantitative Ethnologically, for it, most could not find demonstrable (1970:230-231) support because of incomplete likely ethnographic reporting. Ar for mortuary the information measures chaeologically, keys have been to data from the Buchanan Thomas King Reservoir, (1976) applied by located in the southern Sierra foothills of California. The resulting mea sures indicated highly structures. For comparative tree-like purposes, Saxe

the same measures for keys of Late Woodland King computed mortuary from the Illinois Valley systems 1975b). The results of this (Tainter are of uncertain cases for in both archaeological significance, application the keys were produced to mini statistical through techniques designed the degree of entropy in the resulting classification. mize The two latter hypothesis is complex It uses (Saxe 1970:112-118). scales: the Depth of Persona (C) scale. The first (D) scale and the Contrast

130

JOSEPH A. TAINTER

in a given disposal attributes type. For as the ratio of the number of positive any disposal type in the disposal mode in a disposal mode attributes (n) over the number that has the highest number (nmax). Thus, D= n/nmax. The C scale is based on the number of contrast sets necessary to define a disposal type. In this n contrast to define a given sets are required case, C= n/nmax, where is based on the number of positive it is measured and nmSLX the maximum burial mode, is number required by any disposal type in the domain. A disposal is characterized type that is unique for all its components by set to differentiate It will require only one contrast maximal redundancy. it. For reasons already discussed, such a disposal type would signify a In this case, the lower persona of high social significance. C scale, the more discrete will be the burial type. On the other hand, a burial mode that requires many it will reflect a social persona differentiate sharing many in such Social distinctiveness istics with other personae. social status. In an reflecting insignificant continuously, the value contrast on the sets to

social character a case is graded soci egalitarian

between the C ety such continuous grading will cause a linear relationship so will As the social significance and/) scales. of the deceased increases, the number of positive in the disposal domain, and so also will attributes sets necessary to define such a disposal mode. the number of contrast In contrast, more complex societies will display graded, rather than levels of ranking. In such societies mortuary will continuous, practices to persons of distinctive exhibit multiple, redundant attributes restricted status. Given sets necessary to such redundancy, the number of contrast classes will be lower. Such societies will have mortuary define disposal in its domains characterized by several social strata, each of which may, show a linear relationship between the C and D scales. domain, mortuary strata the relationship will not be linear, since in each between However, higher stratum the number of contrast modes will be lower than in the previous provides to differentiate sets needed burial a measure lower stratum. Hence, the C and D scales in the relationship between system. (Saxe

of the degree of linearity a quantification of the social of the degree of complexity this hypothesis of proved promising testing Ethnographic but it has not been applied archaeologically. 1970:231-233), Another

the to quantitative utilizes analysis of past societies approach of systems theory (Tainter 1975b, 1977a, b; Tainter and Cordy perspective in a systems perspective societies suggests that 1977). Viewing prehistoric our goal is to model entities. Such a such societies as complex, integrated goal entails information dimensions a considerable for which shift in the types of social dimensions on familiar studies concentrate is sought. Most archaeological or ranking. A consideration of prehistoric such as residence

PREHISTORIC

MORTUARY

PRACTICES

131

as systems A requires a shift to a higher level of abstraction. cannot be characterized the attributes system simply merely by describing of its parts. A system is a holistic entity the characteristics of which are not equivalent to the summed characteristics its individual parts (von of a social system we must isolate and 1968:55). To characterize Bertalanffy measure social dimensions that reflect the constituent parts of a system, societies as the patterned among these parts. Such abstract, relationships are the structure and the organization dimensions of social composite to indicate the number, The structure of a system is meant systems. and subsystems of its articulated nature, and arrangement components is most basically defined as (Miller 1965:209, 218), whereas organization on the ranges of behavior which may possibly the constraints be imposed the elements of a system (Rothstein 1958:34-36). pursued by This perspective for which information suggests four social dimensions are the degree of structural differentiation, is to be sought. These the nature and of structural differentiation, the amount or degree of organization, the nature of organization. The degree of structural differentiation are quantitative as and must be modeled and of organization variables,
such.

as well

For archaeological of structural analysis, we can isolate two dimensions that are of general significance. differentiation These might be termed the vertical and the horizontal dimensions. The first refers to the structure of rank grading in a society. The horizontal struc dimension encompasses tural components that equate on identical hierarchical levels. Examples of the latter might include sodalities, certain descent types of coresident territorial bands, and the like. units, task groups, we In terms of structural differentiation along the horizontal dimension, have seen how the spatial organization of mortuary remains can reveal the existence of corporate descent groups. Yet there are many components of the horizontal for which such universal dimension criteria for analysis have not been developed. The study of the vertical dimension presents fewer ambiguities. of ranking may be confidently Levels isolated as dis tinctive levels of energy expenditure (Tainter 1975b; cf. Randsborg 1974). As Blau (1970:203-204) has noted, such levels of ranking may be viewed as structural components of a social system. The number of such rank levels will mark dimension. The analysis of vertical differentiation allows the indirect measurement of the total structural complexity In an analysis of a social system. of on contemporary structural differentiation Blau (1970) organizations found that the size of an organization correlated strongly with both its and its degree of horizontal differentia degree of vertical differentiation the degree of structural differentiation along the vertical

132

JOSEPH A. TAINTER

is not limited by the function this finding of an can be considered to the observation Hence, applicable organization. as well as to contemporary social systems. This finding is prehistoric cor for archaeological research. The strength of the observed significant of a relations (see Blau 1970) suggests that any one of these characteristics tion. Blau noted that not only itself, but the state of the system may be used to monitor as well. Structural other two differentiation along the vertical dimension can thus serve as an index of the total structural complexity of past social social
systems.

is a more complex matter. Organiza on behavior (Rothstein 1958:34-36). must be approached of organization through the measurement The field that provides of behavioral constraint. techniques is information of organization for the measurement theory. of information The central concepts theory that are of use to us are The measurement tion has been defined the measurement Thus, of organization as constraint or and its converse, randomness), (disorder, disorganization, is statistical In communi An additional concept ganization. equilibrium. in a set of messages, cation theory this would refer to a situation where, A situation of all messages of transmission. have an equal probability to a condition of entropy. In statistical statistical equilibrium corresponds entropy and of any specific message is unpredictable, the occurrence equilibrium thus the maximum information may be obtained when a particular mes that covary and entropy are quantities information sage is sent. Hence,

(Weaver 1949;Goldman 1953;Theil 1967;Young 1971).

some use of these concepts, consider the archaeological a number of into differentiated social system structurally hypothetical In the total absence the constraints of any organizational components. in the components individuals process acquire membership by which result in a would be purely random. Such a random process will ultimately of the population the proportion situation where selecting affiliation with a fraction equal to UN, where N is the number would be each component To illustrate to a situa Such a proportional distribution of components. corresponds and so to a condition of entropy. tion of statistical equilibrium, of one form or lacks constraints But of course no social situation limit another. To the extent that either social or demographic pressures as those of to acquire membership in certain components freedom (such of In the presence and thus organization. high status), there is constraint with certain of the population the proportion such constraints, affiliating It is will depart from a condition of statistical components equilibrium. that the mathematical tech from equilibrium such departures precisely 1975b, (Tainter theory may be used to measure niques of information 1977a, b; Tainter and Cordy 1977).

PREHISTORIC

MORTUARY

PRACTICES

133

Shannon's transformed given as

(1949:50-51) in a number

formula of ways. N

for measuring For our purposes

information

may the formula may

be be

H=
where H

^Pilog2 i=l

l/pi

of a set of probabilities, is the entropy (information) and/?* is the of the ith message of occurrence (or the ith structural compo probability to log2 N ranges from 0.0 (no entropy) nent). This information measure en are equally probable or components when all messages (maximum referred to as Hmax). tropy, as the that organization be measured Rothstein (1958:36) has suggested excess of the entropy a system may potentially less the exhibit (Hmax), refers to this exhibit (1972:35-36) (H). Gatlin from Equiprobability, asDj. The symbolized = can range from 0.0 (maximum formula for organization H) (D1 Hmax on //max (maximum organization). to a value that converges entropy) that the amount of in the calculation of the D1 measure It is apparent a system may potentially exhibit increases with its degree of organization a measure Since structural complexity. o? amount of organization. Dx is it actually does as the Divergence

entropy measure

inHmax, it is useful to the maximum Dx value may increase with increases o? degree of organization. the ratio RD1 = DJHmax as a measure calculate on a population's access to Where Dx assesses the amount of constraint con how much in structural components, RD1 measures membership straint is imposed relative to the amount that might be imposed. The ratio RD x 0.0 (maximum will range between relative entropy) and a value that 1.0 (maximum relative organization). approaches to be mea Within the final variable the perspective being discussed, an impor is considered Rank differentiation sured is rank differentiation. tant variable to measure and organization. Levels nents of a system, while it includes elements of both structure because of ranking may be viewed as structural compo is clearly an constraint) (behavioral organization element of rank differentiation. A measure is for rank differentiation (1959:23-25) by Harary proposed In this measure useful for mortuary studies. the status of an individual is

to subordinate based on two interacting factors: the number of persons the individual, and the number of rank grades that separate these subordi nates from the person of higher status. In a ranked hierarchy, any person P will downward in rank level k (rank levels are numbered nk subordinates from any given level, so that the first lower level is 1, the second is 2, etc.). IfM is the number of rank levels below P, the status s(P) of person P can be measured by the formula have

134

JOSEPH A. TAINTER

s(P)=? k(nk) k=l

the status measure of a person is the number of the verbally, immediate subordinates, plus twice the number of their im mediate subordinates, plus three times the number of their immediate and so forth. Hence, subordinates, persons at the top of a hierarchy will a higher status measure have if they have a greater number of subordi Expressed individual's nates and/or if their subordinates are heavily distributed in the bottom most rank levels. The rationale behind this measure is discussed in more detail by Harary (1959) and Tainter 1977a, b). (1975b, In archaeological to determine it is impossible the precise analysis to a person of paramount number of individuals subordinate rank at any to determine the number of levels point in time. But it should be possible of ranking in a past social system, and to ascertain the number of persons from a total mortuary of each rank grade. population who were members To compute it is necessary to consider the s(P) measure the population of rank levels as multiples subordinate of the number of per (or fractions) in the paramount rank grade. The formula used for this purpose in in rank level k, and Nx where Nk refers to the number of persons NjJNl9 indicates the number in the highest rank level. With this transformation the paramount rank grade is always associated with the value 1.0, and subordinate levels become appropriate multiples of 1.0. The (or fractions) can then be expressed as the in a hierarchy degree of rank differentiation difference the highest and lowest levels. Since the status of the between level is always can be lowest the degree of rank differentiation zero, as s(l), the status measure of the highest level. expressed that the archaeological of these measures application sets of mortuary data. The use of the measures may be requires complete to a set of Middle Woodland illustrated by reference 150 b.c.-a.D. (ca. bluffs of the Illinois situated on the western 400) burials from mounds In a set of 512 interments, the following levels of energy expendi Valley. ture and ranking were Level Level Level Level Level Level observed (Tainter 1975b, 1977a, b). through, large, log-roofed feature of most mounds 1: Individuals interred in, or processed served as the central tombs, which It should be clear sons

in smaller, burials 2: Peripheral graves log-covered slab inclusions limestone 3: Burials with sociotechnic 4: Individuals buried with locally produced most often Hopewell series pottery vessels 5: 6: Interments Individuals in simple subfloor graves mound placed on aceretional surfaces.

items,

PREHISTORIC

MORTUARY

PRACTICES

135

of The distribution Table 4.1. This table sure for this system. entropy of 1.8354 bits. measure of amount of degree of organization of rank differentiation 5(1) amounts system,

also

among these rank levels is shown in of the entropy mea the calculation an rank system displays The Middle Woodland for six rank levels is 2.58496 bits, and so the Hmax 1.8354) .7496. The (DJ is (2.58496 organization .29. The measure to (.7496/2.58496) (RDj amounts shown in Table 4.2. For this Middle Woodland is to 13.671. the population illustrates

measures are most meaningful when used for the Such quantitative In two such applications assessment these mea of variation and change. sures have yielded excellent social change results. In a study of Woodland a there appeared in west-central Illinois during the period a.D. 200-800 in the structural complexity of social systems between general decrease and Late Woodland the Middle (a.d. 400-900) periods. Also evident are two contrasting in both organization trends: (a) a decrease quantitative in the early Late Woodland and degree) and rank differentiation (amount in these variables increase during later period; and (b) a corresponding to the emergence of the truly complex social Late Woodland, just prior The pattern of social characteristic of the Mississippian systems period. not only for is significant, change evidenced by the quantitative measures but also because it parallels and is confirmed it discloses, assessments 1975b, 1977a, b). It might be (Tainter by nonquantitative had been used in this added at this point that, if evolutionary typologies the information and Late Woodland systems would have been charac study, both Middle no social terized as ranked societies Thus, (in Fried's terminology). the Middle and Late Woodland periods would have been change between evident. The pattern of social change that was observed was evident only measures. because of the use of quantitative to two contem In a second have been applied study these measures TABLE 4.1 Entropy of Middle Woodland Rank level Rank System p
.205 1 .131 2 .025 3 .011 4 .544 5 .084 6

log2 1/p
2.28630 2.93236 5.32193 6.50635 .878321 3.57347

p log2 1/p
.4687 .3841 .1330 .0716 .4778 .3002 1.8354

136

JOSEPH A. TAINTER

TABLE

4.2

Middle Woodland
Rank

Rank Differentiation
level NJN,N 1 91 1.000 58 2 11 3 5 241 5 37 6 4 .637 .121 .055 2.648 .407 s(1) 13.671

social systems from the Island of Hawaii. One of poraneous prehistoric a complex these (Kaloko) rank hierarchy evidenced displays by both data and settlement patterns. The other (Anaehoomalu) mortuary displays a very truncated rank hierarchy with no persons of the highest rank these qualitative conclusions, analysis corroborates grades. Quantitative that Kaloko was characterized and indicates by greater structural com and rank differentiation (Tainter and Cordy 1977). organization, plexity, A comparative that applies the quantitative methods study developed by Saxe (1970) and Tainter (1975b) to the same data sets would be a useful step in furthering mortuary studies.

MORTUARY

PRACTICES:

CONCLUDING

EVALUATIONS

This discussion of the social dimensions of mortuary leads to practices a set of conclusions re and recommendations for future archaeological that investigators search. Perhaps most basic is the recommendation be about which dimensions in of a social system are to be monitored explicit a given study. (It does not need to be emphasized that this suggestion on mortuary to all studies of past societies, whether applies dependent data or not.) Along with the recognition that social dimensions of interest to model should be explicitly stated goes the responsibility these dimen that will allow the study of social variation and change. sions in a manner The minimum entailed by this consideration is that social requirement and cross-culturally dimensions be modeled by criteria that are objective valid. For mortuary two analytical criteria have been developed practices that fulfill these requirements. of mor These are the spatial distribution that contains information tuary remains, a variable relating to corporate an indicator of rank grad and energy expenditure, group differentiation, ing. A second consideration in the study of variation and change is that

PREHISTORIC

MORTUARY

PRACTICES

137

on true interval or ratio scales. The nomi social dimensions be measured are simply of evolutionary that are characteristic nal categories typologies to yield the kinds of information we need for the not sensitive enough to quantitative of past modeling study of social change. Two approaches Tainter social systems have been developed 1975b). Such (Saxe 1970; a central goal variables must become measurement of social quantitative in the study of mortuary practices.

REFERENCES
Autry, W. 1974 Analysis presented Baker, Washington, B. L. 1974 Statistical of Post-Formative at the 1974 Annual D.C. burial Meeting Valley practices: of the Society of Oaxaca, Mexico. Paper for American Archaeology,

of the relative data: A re-evaluation of Maya mortuary analysis at the of males and females at Altar de Sacrificios. Paper presented importance Mexico of the American 1974 Annual Association, Anthropological Meeting

Binford,

City. L. R. 1962 1971 as anthropology. 28:217-225. American Antiquity In Approaches Their study and their potential. practices: of mortuary edited by J. A. Brown. Society practices, Memoirs 25:6-29. Archaeology, Archaeology A formal of differentiation in organizations. American to the social for American

Mortuary dimensions

Blau,

P. M. 1970 theory 35:201-218. Sociological

Review Bowen, R. N. 1961 Hawaiian

disposal

of

the dead. of Hawaii,

thropology, Brown, J. A. 1971 The

University of

Unpublished Honolulu.

M.A.

thesis,

Department

of An

dimensions

status

in the burials edited

dimensions 1973

Archaeology, Cultural evolution Southeastern

of mortuary Memoirs and

practices, 25:92-112.

at Spiro. In Approaches by J. A. Brown. Society Paper presented

to the social for American at the 1973

the rise and

fall of Spiro. Memphis. A

Archaeological in the lower variability Illinois and

Conference, River

J. E. Buikstra, 1972 Hope

well

valley:

regional

approach

biological

dissertation, Department D. I. Bushnell, 1920 Native and forms cemeteries can Ethnology, Clark, G. 1969 A Bulletin No.

mortuary activity. prehistoric Unpublished of Anthropology, of Chicago. University of burial 71. east of the Mississippi. Bureau

to the study of Ph.D.

of Ameri

of burial analysis preliminary Arizona. 35:57-86. The Kiva

clusters

at the Grasshopper

site,

east-central

138

JOSEPH A. TAINTER

Clarke,

D. 1968

L. archaeology. of the dead London: in western 93:71-97. and problems. Los Angeles) Methuen. Australia. the American

Analytical D. S. Davidson, 1949 Disposal Decker, D. 1969 Philosophical A. Early Survey Ellis, W.

Proceedings

of

Society

on Catalina Island: Potential archaeology Annual Report (University of California,

Archaeological 11:69-84.

1969 Polynesian Hawaii. researches: Tuttle. Tokyo: N. Leonard, C. King, and L. King P., D. Decker, III, T. King, Finnerty, on structure A salvage 1970 Community and trade at Isthmus Cove: excavation Coast Island. Occasional Catalina Society Archaeological Pacific Papers No. Fredrickson, 1974 D. Social Indian Ballena Fried, M. 1967 H. The York: Gatlin, L. A. 1972 Information Press. theory and the living system. New York: Columbia University evolution Random of political House. society, an essay in political anthropology. New 1. A. change in prehistory: and economic A central California In ? Antap:California example. L. Bean and T. King. edited by 2, 57-73.

political Press Anthropological

organization, No. Papers

S. Goldman, 1953 Information L. Goldstein, 1976 Spatial pian Goodenough, 1965

theory. and

New social

York:

Dover. of Mississip Arbor: Univer

structure

' ' ' a general model "status' and of the cultural organi 'role' :Toward Rethinking zation of social relationships. In The relevance for social anthropol of models No. A.S.A., 1, 1-24. ogy, edited by M. Banton. Monographs Status and contrastatus. Sociometry mortuary Pennsylvania social 22:23-43. M.A. University De

Ph.D. society. sity Microfilms. W. H.

dissertation,

manifestations Regional organization: Ann Northwestern University.

Harary, Hatch,

F. 1959 J. W. 1974 Social partment dimensions of Dallas patterns. State Unpublished University, thesis, Park.

of Anthropology, K Pueblo: Prehistoric

Hill,

J. N. 1970 Broken organization Papers No. in the American 18. In Borneo Journal and Southwest.

Hudson,

University A. B. 1966 Death related 416. L. 1969

of Arizona

Anthropological

ceremonies matters,

edited

of the Padju Epat Ma'anyan Dayaks. Sarawak Museum by T. Harrison. No. 1). An

writing

13(27):341

(Special Monograph

King,

Creek (LAn-243): cemetery from mortuary practices. Archaeological 11:23-68. Los Angeles) California,

The Medea

Survey

investigation Annual

of social Report

organization (University of

PREHISTORIC

MORTUARY

PRACTICES

139

King,

T.

F. The San sional dead at Tibur?n: Mortuary customs and social organization on northern Society Bay. Occa

1970

Francisco Papers evolution

Bay. Northwestern No. 2. of status

California

Archaeological

1974

The

1976

political Press Anthropological Papers King. Ballena Political differentiation among hunter-gatherers: dissertation, crofilms. University of California,

California

Indian

around San Francisco ascription edited and economic organization, No. An 2, 37-53.

In ? Antap: and T.

by L. Bean

Riverside.

test. Ph.D. archaeological Ann Arbor: Mi University

Kroeber, Larson,

A. 1927

L. of the dead. American Anthropologist stratification 29:308-315. at the Etowah practices, 25:58-67.

Disposal L. H.

1971

Archaeological In Approaches

implications to the social

of social dimensions

Brown. Society for American W. Longacre, as anthropology: 1970 Archaeology No. 17. thropological Papers McWilliam, 1936 N. D. of the dead

of mortuary Memoirs Archaeology, A case study.

edited

site, Georgia. by J. A.

University

of Arizona

An

Disposal

Mason,

Territory J. A. 1912 D. 1965

of New

Guinea.

in the Buang Mankind

Mountains, 2(2): 39-42.

Morobe

District,

Mandated

The ethnology Indians. of the Salinan American and Ethnology Archaeology Socio-economic of

University 10(4):97-240. burial.

of California

Publications

in

Miles,

aspects Basic

secondary

Oceania Science

35:161-174. 10:193-237.

Miller, Peebles,

G. 1965 C. 1971 Living S. systems: and concepts. Behavioral

Moundville practices, edited by 91.

of mortuary considerations structural sites: Some surrounding to the social dimensions In Approaches of mortuary practices, Memoirs 35:68 J. A. Brown. Society for American Archaeology, II. of the Moundville 8(2): 1-13. of a prehistoric Santa burials: An initial report. News and culture. Ph.D.

1972 1974

Monothetic-divisive of Computer The Moundville: dissertation, Microfilms. letter

analysis Archaeology

University

organization of California,

community Ann Barbara.

Arbor:

University

Randsborg, 1974 1975 W. 1970

K. Social cultural Social Society L. stratification systems. dimensions 41:105-118. A study in early Bronze Age Denmark: Praehistorische 49:38-61. Zeitschrift Denmark. of early Neolithic Proceedings in the regulation of the Prehistoric of

Rathje,

Sociopolitical tive hypotheses. Models change: Models 731-757.

of Lowland implications World Archaeology

burials: Maya 1:359-374.

Methodology

and

tenta

1973

for mobile

In The explanation Maya: A variety of constraints. of culture in prehistory, London: edited by C. Renfrew. Duckworth. Pp.

140

JOSEPH A. TAINTER

Ray,

V. 1963 Primitive The Modoc pragmatists: Press. of Washington social Indians of Northern California. Seattle:

Service,

University E. R. 1962 1971 Primitive House. Cultural

organization, Theory

an evolutionary in practice.

perspective. New York:

New Holt.

York:

Random

evolutionism:

C. E. Shannon, 1949 The

mathematical Press.

communication, Illinois S. Shennan, 1975 The Renfrew, 1973 C. Monuments,

of communication. theory and W. by C. E. Shannon Pp. 29-125. at Branc. Antiquity

In The mathematical Weaver. Urbana:

theory University

of of

social

organization

49:279-288.

1974

In The in neolithic Wessex. and social organization mobilization, in prehistory, edited by C. Renfrew. of culture change: Models explanation London: Duckworth. Pp. 539-558. a subsistence in prehis of social organization The evolution economy: Beyond Bulle In Reconstructing edited by C. Moore. toric Europe. societies, complex tin of the American Schools headhunters. of burials Research Notes of Oriental Research No. 20. 69-85. (Supplement)

Riley,

E. Philadelphia: in the lower Lippincott. Snake River region. Northwest An

1925 Among Papuan M. J. Rodeffer, 1973 A classification thropological N. A.

7:101-131. at New the 1973 Annual of the

Rothschild, 1973

Rothstein, 1958 M. 1968 Saxe, A. 1970 A.

and status. Sex, graves, American Anthropological J. Communication, Press. D. Tribesmen. Social Michigan. Englewood

Paper

presented

Meeting

Association, and

Orleans. Indian Hills, Colo.: Falcon's

organization,

science.

Wing

Sahlins,

Cliffs.

N.J.:

Prentice-Hall. of

dimensions Ann

of mortuary Arbor: University

Ph.D. practices. Microfilms. site. Archaeological 10:209-261.

dissertation,

University

Stickel,

E. G. 1968 Status differentiations of California, excavations northern at the Rinc?n Los Survey Annual Report

Tainter,

(University J. A. 1971 Salvage of the

Angeles)

at the Fowler Chumash. San

site: Some Luis Obispo

aspects County

of the social Archaeological

organization Society Hawaii.

Occasional 1973 1974 The social

Archaeology

3. No. Papers North at Kaloko, of mortuary correlates patterning 8:1-11. in Oceania and Physical Anthropology

Kona,

to the archaeological change. Depart study of social approaches Analytical Illinois ment of Anthropology, Northwestern Evanston, (Xerox). University, classifica in numerical An experiment and mortuary inference 1975a Social practices: 7:1-15. tion. World Archaeology 1975b The archaeological study of social change: Woodland systems in west-central

PREHISTORIC

MORTUARY

PRACTICES

141

Illinois. Microfilms. 1976 Spatial Hawaii.

Ph.D.

dissertation,

Northwestern

University.

Ann

Arbor:

University

organisation Archaeology change edited

1977a Modeling chaeology, 1977b Woodland

in the Kaloko North Kona, cemetrery, patterning 11:91-105. and Physical in Oceania Anthropology In For in ar in prehistoric social systems. theory building Press. New York: Academic Pp. 327-351. by L. R. Binford. and social in west-central Illinois. Mid-Continental Journal of

social change 2:67-98. Archaeology J. A., and R. H. Cordy Tainter, 1977 An archaeological analysis Hawaii. Theil, H. 1967 Economics and information World Archaeology

of social 9:95-112.

ranking

and residence

groups

in prehistoric

theory.

Amsterdam:

North-Holland

Publishing

Co.

D. H. Ubelaker, 1974 Reconstruction study from 18. No. Ucko, P. 1969 S. 1975 Sociocultural practices. University von L. Bertalanffy, 1968 General W. 1949 Recent Urbana: Whallon, 1972 Winters, H. 1968 Value R. A new J. Ethnography chaeology Vehik,

of demographic from profiles Potomac. Smithsonian the tidewater

skeletal ossuary Contributions

samples: A case to Anthropology

and archaeological 1:262-280. of

interpretation

of funerary

remains.

World

Ar

implications Ph.D. dissertation, Microfilms. New

central University

early European of Missouri,

Bronze Columbia.

Age mortuary Ann Arbor:

system

theory.

York:

Braziller. In The Weaver.

Weaver,

contributions theory

mathematical

University approach systems

to the mathematical theory of communication. and W. of communication, by C. E. Shannon 1-28. of Illinois Press. Pp. typology. American Antiquity 37:13-33.

to pottery and trade

perspectives Aldine. Pp. Young, J. F. 1971 Information

in archeology, 175-222. New

in the Late Archaic in the Midwest. cycles edited and L. Binford. by S. Binford

In New Chicago:

theory.

York:

Wiley

(Interscience).