Survey Research Methods 5th Edition Floyd J Fowler
1. Survey Research Methods 5th Edition Floyd J
Fowler download
https://ebookbell.com/product/survey-research-methods-5th-
edition-floyd-j-fowler-6846390
Explore and download more ebooks at ebookbell.com
2. Here are some recommended products that we believe you will be
interested in. You can click the link to download.
Encyclopedia Of Survey Research Methods 2 Volume Set 1st Edition Dr
Paul J Lavrakas
https://ebookbell.com/product/encyclopedia-of-survey-research-
methods-2-volume-set-1st-edition-dr-paul-j-lavrakas-1377624
Doing Survey Research Guide To Quantitative Methods Peter M Nardi
https://ebookbell.com/product/doing-survey-research-guide-to-
quantitative-methods-peter-m-nardi-46223598
Doing Survey Research A Guide To Quantitative Methods 4th Edition
Peter M Nardi
https://ebookbell.com/product/doing-survey-research-a-guide-to-
quantitative-methods-4th-edition-peter-m-nardi-48147650
Experimental Methods In Survey Research Mick P Couper Graham Kalton
https://ebookbell.com/product/experimental-methods-in-survey-research-
mick-p-couper-graham-kalton-10679720
3. Small Populations Large Effects Improving The Measurement Of The Group
Quarters Population In The American Community Survey 1st Edition
National Research Council Division Of Behavioral And Social Sciences
And Education Committee On National Statistics Panel On Statistical
Methods For Measuring The Group Quarters Population In The American
Community Survey Krisztina Marton Paul R Vos
https://ebookbell.com/product/small-populations-large-effects-
improving-the-measurement-of-the-group-quarters-population-in-the-
american-community-survey-1st-edition-national-research-council-
division-of-behavioral-and-social-sciences-and-education-committee-on-
national-statistics-panel-on-statistical-methods-for-measuring-the-
group-quarters-population-in-the-american-community-survey-krisztina-
marton-paul-r-vos-51980318
Research Methods In Community Medicine Surveys Epidemiological
Research Programme Evaluation Clinical Trials Sixth Edition J H
Abramson
https://ebookbell.com/product/research-methods-in-community-medicine-
surveys-epidemiological-research-programme-evaluation-clinical-trials-
sixth-edition-j-h-abramson-4311502
Landscape And Interaction The Troodos Archaeological And Environmental
Survey Project Cyprus Volume 1 Methodology Analysis And Interpretation
Michael Given
https://ebookbell.com/product/landscape-and-interaction-the-troodos-
archaeological-and-environmental-survey-project-cyprus-
volume-1-methodology-analysis-and-interpretation-michael-
given-48874326
Survey Research And Sampling Jan Eichhorn
https://ebookbell.com/product/survey-research-and-sampling-jan-
eichhorn-47176688
Survey Research In The United States Roots And Emergence 18901960 1st
Edition Jean M Converse
https://ebookbell.com/product/survey-research-in-the-united-states-
roots-and-emergence-18901960-1st-edition-jean-m-converse-33178612
7. APPLIED SOCIAL RESEARCH
METHODS SERIES
1. SURVEY RESEARCH METHODS (Fifth Edition)
b y FLOYD J. FOWLER, Jr.
2. SYNTHESIZING RESEARCH (Fourth Edition)
b y HARRIS COOPER
3. METHODS FOR POLICY RESEARCH (Second Edition)
b y ANN MAJCHRZAK
4. SECONDARY RESEARCH (Second Edition)
b y DAVID W. STEWART and MICHAEL A. KAMINS
5. CASE STUDY RESEARCH (Fifth Edition)
b y ROBERT K. YIN
6. META-ANALYTIC PROCEDURES FOR SOCIAL
RESEARCH (Revised Edition)
b y ROBERT ROSENTHAL
7. TELEPHONE SURVEY METHODS (Second Edition)
b y PAUL J. LAVRAKAS
8. DIAGNOSING ORGANIZATIONS (Second Edition)
b y MICHAEL I. HARRISON
9. GROUP TECHNIQUES FOR
IDEA BUILDING (Second Edition)
b y CARL M. MOORE
10. NEED ANALYSIS
b y JACK McKILLIP
11. LINKING AUDITING AND META EVALUATION
b y THOMAS A. SCHWANDT
and EDWARD S. HALPERN
12. ETHICS AND VALUES
IN APPLIED SOCIAL RESEARCH
b y ALLAN J. KIMMEL
13. ON TIME AND METHOD
b y JANICE R. KELLY
and JOSEPH E. McGRATH
14. RESEARCH IN HEALTH CARE SETTINGS
b y KATHLEEN E. GRADY
and BARBARA STRUDLER WALLSTON
15. PARTICIPANT OBSERVATION
b y DANNY L. JORGENSEN
16. INTERPRETIVE INTERACTIONISM (Second Edition)
b y NORMAN K. DENZIN
17. ETHNOGRAPHY (Third Edition)
b y DAVID M. FETTERMAN
18. STANDARDIZED SURVEY INTERVIEWING
b y FLOYD J. FOWLER, Jr.,
and THOMAS W. MANGIONE
19. PRODUCTIVITY MEASUREMENT
b y ROBERT O. BRINKERHOFF
and DENNIS E. DRESSLER
20. FOCUS GROUPS (Second Edition)
b y DAVID W. STEWART,
PREM N. SHAMDASANI, and DENNIS W. ROOK
21. PRACTICAL SAMPLING
b y GART T. HENRY
22. DECISION RESEARCH
b y JOHN S. CARROLL
and ERIC J. JOHNSON
23. RESEARCH WITH HISPANIC POPULATIONS
b y GERARDO MARIN
and BARBARA VANOSS MARIN
24. INTERNAL EVALUATION
b y ARNOLD J. LOVE
25. COMPUTER SIMULATION APPLICATIONS
b y MARCIA LYNN WHICKER and LEE SIGELMAN
26. SCALE DEVELOPMENT (Third Edition)
b y ROBERT F. DeVELLIS
27. STUDYING FAMILIES
b y ANNE P. COPELAND and KATHLEEN M. WHITE
28. EVENT HISTORY ANALYSIS
b y KAZUO YAMAGUCHI
29. RESEARCH IN EDUCATIONAL SETTINGS
b y GEOFFREY MARUYAMA
and STANLEY DENO
30. RESEARCHING PERSONS WITH
MENTAL ILLNESS
b y ROSALIND J. DWORKIN
31. PLANNING ETHICALLY
RESPONSIBLE RESEARCH (Second Edition)
b y JOAN E. SIEBER and MARTIN TOLICH
32. APPLIED RESEARCH DESIGN
b y TERRY E. HEDRICK,
LEONARD BICKMAN, and DEBRA J. ROG
33. DOING URBAN RESEARCH
b y GREGORY D. ANDRANOVICH
and GERRY RIPOSA
34. APPLICATIONS OF CASE STUDY RESEARCH (Third Edition)
b y ROBERT K. YIN
35. INTRODUCTION TO FACET THEORY
b y SAMUEL SHYE and DOV ELIZUR
with MICHAEL HOFFMAN
36. GRAPHING DATA
b y GARY T. HENRY
37. RESEARCH METHODS IN SPECIAL EDUCATION
b y DONNA M. MERTENS
and JOHN A. McLAUGHLIN
38. IMPROVING SURVEY QUESTIONS
b y FLOYD J. FOWLER, Jr.
39. DATA COLLECTION AND MANAGEMENT
b y MAGDA STOUTHAMER-LOEBER
and WELMOET BOK VAN KAMMEN
40. MAIL SURVEYS
b y THOMAS W. MANGIONE
41. QUALITATIVE RESEARCH DESIGN (THIRD EDITION)
b y JOSEPH A. MAXWELL
42. ANALYZING COSTS, PROCEDURES,
PROCESSES, AND OUTCOMES
IN HUMAN SERVICES
b y BRIAN T. YATES
43. DOING LEGAL RESEARCH
b y ROBERT A. MORRIS, BRUCE D. SALES,
and DANIEL W. SHUMAN
44. RANDOMIZED EXPERIMENTS FOR PLANNING
AND EVALUATION
b y ROBERT F. BORUCH
45. MEASURING COMMUNITY INDICATORS
b y PAUL J. GRUENEWALD, ANDREW J. TRENO,
GAIL TAFF, and MICHAEL KLITZNER
46. MIXED METHODOLOGY
b y ABBAS TASHAKKORI and CHARLES TEDDLIE
47. NARRATIVE RESEARCH
b y AMIA LIEBLICH, RIVKA TUVAL-MASHIACH, and
TAMAR ZILBER
48. COMMUNICATING SOCIAL SCIENCE RESEARCH
TO POLICY-MAKERS
b y ROGER VAUGHAN and TERRY F. BUSS
49. PRACTICAL META-ANALYSIS
b y MARK W. LIPSEY and DAVID B. WILSON
50. CONCEPT MAPPING FOR PLANNING
AND EVALUATION
b y MARY KANE and
WILLIAM M. K. TROCHIM
51. CONFIGURATIONAL COMPARATIVE METHODS
b y BENOÎT RIHOUX and CHARLES C. RAGIN
8. Survey Research Methods
Fifth Edition
Floyd J. Fowler, Jr.
Center for Survey Research,
University of Massachusetts Boston
10. Contents
Preface ix
About the Author xi
1. Introduction 1
2. Types of Error in Surveys 8
3. Sampling 14
4. Nonresponse: Implementing a Sample Design 42
5. Methods of Data Collection 61
6. Designing Questions to Be Good Measures 75
7. Evaluating Survey Questions and Instruments 99
8. Survey Interviewing 110
9. Preparing Survey Data for Analysis 127
10. Analyzing Survey Data 134
11. Ethical Issues in Survey Research 140
12. Providing Information About Survey Methods 146
13. Survey Error in Perspective 150
References 155
Author Index 164
Subject Index 168
11. Detailed Contents
Preface ix
About the Author xi
1. Introduction 1
Reasons for Surveys 1
Components of Surveys 3
Purposes and Goals of This Text 7
2. Types of Error in Surveys 8
Error Associated With Who Answers 9
Error Associated With Answers 11
Recapping the Nature of Error in Surveys 12
3. Sampling 14
The Sample Frame 15
Selecting a One-Stage Sample 18
Multistage Sampling 21
Drawing Samples From Two or More Sample Frames 30
Making Estimates From Samples and Sampling Errors 32
How Big Should a Sample Be? 37
Sampling Error as a Component of Total Survey Error 39
4. Nonresponse: Implementing a Sample Design 42
Calculating Response Rates 43
Bias Associated With Nonresponse 43
Reducing Nonresponse 49
Two Other Approaches to Reducing Nonresponse Error 54
Nonprobability (or Modified Probability) Samples 55
Nonresponse as a Source of Error 58
5. Methods of Data Collection 61
Major Issues in Choosing a Strategy 61
Summary Comparison of Methods 71
Conclusion 73
6. Designing Questions to Be Good Measures 75
Increasing the Reliability of Answers 76
Types of Measures/Types of Questions 86
12. Increasing the Validity of Factual Reporting 91
Increasing the Validity of Answers Describing Subjective States 96
Question Design and Error 97
7. Evaluating Survey Questions and Instruments 99
Defining Objectives 100
Preliminary Question Design Steps 100
Presurvey Evaluation 102
Design, Format, and Layout of Survey Instruments 104
Field Pretests 105
Survey Instrument Length 108
Conclusion 109
8. Survey Interviewing 110
Overview of Interviewer Job 110
Interviewer Recruitment and Selection 113
Training Interviewers 115
Supervision 117
Survey Questions 120
Interviewing Procedures 120
Validation of Interviews 122
The Role of Interviewing in Survey Error 123
9. Preparing Survey Data for Analysis 127
Formatting a Data File 127
Constructing a Code 128
Approaches to Coding and Data Entry 129
Data Cleaning 132
Coding, Data Entry, and File Creation as Sources of Errors 132
10. Analyzing Survey Data 134
Adjusting for Sample Nonresponse and Sample Frame Deficiencies 134
Coping With Item Nonresponse 136
Adjusting for Different Probabilities of Selection 137
Calculating Sampling Errors 138
Conclusion 139
11. Ethical Issues in Survey Research 140
Informing Respondents 141
Protecting Respondents 142
Benefits to Respondents 144
Ethical Responsibilities to Interviewers 144
Conclusion 145
13. 12. Providing Information About Survey Methods 146
13. Survey Error in Perspective 150
The Concept of Total Survey Design 150
Error in Perspective 150
Conclusion 153
References 155
Author Index 164
Subject Index 168
14. ix
Preface
The goal of this fifth edition of Survey Research Methods, like that of its predecessors,
is to produce a summary of the basic concepts and current knowledge about sources of
error in surveys for those who are not primarily statisticians or methodologists. Surveys
are fundamentally a matter of asking a sample of people from a population a set of
questions and using the answers to describe that population. How the sample is
selected, which questions are asked, and the procedures used to collect the answers all
have the potential to affect how well the survey is likely to accomplish its goals. If one
is going to commission a survey or use survey data collected by others, it is important
to understand why these issues matter and how they affect survey results. Readers
should have that understanding by the time they finish this book.
Considerable effort has been made to make this book accessible to a general audi-
ence. Although familiarity with social science research and statistical concepts is a plus,
no special background should be required to grasp the material in this book.
This is also designed to be a comparatively brief book. Choices have been made
about the level of depth given to the various topics. Throughout the book, there are
suggestions for further reading for those whose interests go beyond an introductory
level.
NEW IN THE FIFTH EDITION
In the past decade or so, there are two profound changes that have been going on in the
survey research world. One change is the growing challenge of collecting data about
the general population by telephone survey. Driven by the increased use of cell phones
and the declining rates at which people respond to telephone requests to do surveys, the
reliance on random-digit dialing telephone samples as a way of doing general popula-
tion surveys is declining. Those who still use this approach are finding it harder and
harder to meet traditional standards for response rates. At the same time, there is a
major effort to try to figure out what the best alternatives are. Technology, in the form
of ever-growing access to the Internet, smart phones, and Interactive Voice Response
(IVR) provide researchers with new options for how to collect data. Sampling addresses
has also become easier as better lists of addresses have become available, leading to
another look at the value of mail surveys. How best to use these resources, singly or in
combination, to collect high-quality data is a work in progress. Change is inevitable,
but a major challenge of this edition was to put these issues in perspective, even as we
know that practices will continue to evolve.
In addition, of course, this edition integrates new studies and publications from the
5 years since the 4th edition was published. Keeping the information current is one of
the main reasons for creating a new edition. However, as I was revising the book, I was
struck by the number of issues for which the best, more informative studies were done
15. x SURVEY RESEARCH METHODS
well in the past. So, when a reference from, say, the 1970s is used, readers should not
think that the information is out of date. Most likely it is still one of the best sources of
information about a particular issue.
ACKNOWLEDGMENTS
Doing justice to the people who have contributed to this book gets harder with each
edition, as the list inevitably grows. I think it is still appropriate to start with the three
people who probably had the most effect on my understanding of survey research meth-
ods: Robert Kahn, Morris Axelrod, and Charles Cannell. In many respects, the task of
the book is to pull together and summarize what others have written and learned, so the
references and, in particular, those suggested for further reading were key resources.
However, the name of Robert Groves is probably found as often as any other in this
edition, and that certainly reflects his large and varied contributions to the field of sur-
vey research.
I would like to specifically thank Tony Roman, Mary Ellen Colten, Trish Gallagher,
Carol Cosenza, and Dragana Bolcic-Jankovic at the Center for Survey Research for
their reviews and helpful comments on various chapters. The Center for Survey
Research provided critical support services. Five reviewers kindly provided feedback
on the 4th edition that helped shape these revisions. They were Joseph C. Kush,
Duquesne University; Gilbert A. Jacobs, Mercyhurst University; Claudette M. Peterson,
North Dakota State University; Candan Duran-Aydintug, University of Colorado
Denver; and Karen A. Thornton, Barry University. Finally, Judy Chambliss, as always,
played a crucial role in helping me to maintain the mental health this effort required.
I thank these and others for their valuable contributions, but, of course, the responsi-
bility for the final product, good and bad, is basically mine.
Jack Fowler
16. xi
About the Author
Floyd J. Fowler, Jr., is a graduate of Wesleyan University and received a PhD from the
University of Michigan in 1966. He has been a Senior Research Fellow at the Center
for Survey Research at UMass Boston since 1971. He was Director of the Center for
14 years. In addition to this book, Dr. Fowler is the author (or co-author) of three other
textbooks on survey methods, including Improving Survey Questions, Standardized
Survey Interviewing (with Mangione), and Survey Methodology (with Groves, Couper,
Lepkowski, Singer, & Tourangeau), as well as numerous research papers and mono-
graphs. His recent work has focused on studies of question design and evaluation tech-
niques and on applying survey methods to studies of medical care. In 2013, Dr. Fowler
received the American Association for Public Opinion Research (AAPOR) Award for
Exceptionally Distinguished Achievement.
18. 1
This book is about standards and practical procedures for surveys designed to
provide statistical descriptions of people by asking questions, usually of a sample.
Surveys meld sampling, question design, and data collection methodologies. Those
who want to collect, analyze, or read about survey data will learn how details of
each aspect of a survey can affect its precision, accuracy, and credibility.
The subject of this book is data collection in social surveys. It includes common
procedures, standards for good practice, and the implications of various design deci-
sions for the quality of survey data. The purpose of the book is to give a sound basis for
evaluating data collection procedures to those who would collect, analyze, or read
about survey data. Readers will come to understand the ways in which the details of
data collection are related to the confidence they can have in figures and statistics based
on surveys.
There are many data collection and measurement processes that are called surveys.
This book focuses on those surveys that have the following characteristics:
• The purpose of the survey is to produce statistics, that is, quantitative or numerical descrip-
tions about some aspects of the study population.
• The main way of collecting information is by asking people questions; their answers con-
stitute the data to be analyzed.
• Generally, information is collected about only a fraction of the population, that is, a sample,
rather than from every member of the population.
REASONS FOR SURVEYS
In the U.S. Constitution, it is specified that a survey meeting the previously mentioned
criteria must be carried out every 10 years. In the decennial census, statistics are pro-
duced about a population by asking people questions. No sampling, though, is involved;
data are supposed to be collected about every person in the population.
The purpose of the decennial census is to count people as a basis for ensuring appro-
priate representation in the House of Representatives. As part of the census, it gathers
information about age, how household members are related to one another, and ethnic
1
Introduction
19. 2 SURVEY RESEARCH METHODS
background. However, those data only begin to meet the information needs about the
population. To provide data to fill those information gaps, special-purpose surveys have
become a prevalent part of American life since the 1930s.
Most people are familiar with three uses of survey techniques: the measurement of
public opinion for newspaper and magazine articles, the measurement of political per-
ceptions and opinions to help political candidates in elections, and market research
designed to understand consumer preferences and interests. Each of these well-developed
programs of survey research is aimed primarily at tapping the subjective feelings of the
public. There are, in addition, numerous facts about the behaviors and situations of
people that can be obtained only by asking a sample of people about themselves. There
is probably no area of public policy to which survey research methodology has not been
applied. The following is an abbreviated list of some of the major applications:
• Unemployment rates, as routinely released by the Bureau of Labor Statistics, as well as
many other statistics about jobs and work, are based on household surveys (Current
Population Surveys) carried out by the Bureau of the Census. Parallel surveys of businesses
and industries are carried out to describe production and labor force needs.
• People’s incomes and the way they spend their money constitute another area in which only
surveys can provide reliable data. Patterns of consumer expenditures and their expectations
have proven to be important predictors of trends in the economy.
• The National Health Interview Survey has been carried out by the Bureau of the Census for
the National Center for Health Statistics since the late 1950s. This survey collects basic data
about health conditions, use of health services, and behaviors that affect the risk of illness.
These are all topics about which only good survey research can provide adequate data.
• The main source of data about criminal events traditionally has been police department
records. Police records, however, only include events that people report to the police. For
most crimes involving victims, surveys provide more reliable measures of the rates at which
crimes occur and the characteristics of the victims. The National Crime Survey was
launched in the 1970s to provide such figures. In addition, surveys are the only way to
measure people’s concerns and fears about crime.
• One of the oldest applications of surveys is by the U.S. Department of Agriculture. The
department surveys farmers to estimate the rate at which different crops will be planted and
to predict the availability of various food products.
• Mental health, transportation needs and patterns of use, political behavior, characteristics of
housing (such as its cost and appropriateness to familial needs), and worker satisfaction are
other examples of areas where survey research is used extensively. The largest collector of
survey data in the United States is undoubtedly the federal government, particularly the
Bureau of the Census and the Department of Agriculture. In addition, thousands of individual
surveys are done each year by university, nonprofit, and for-profit survey organizations.
Sponsoring a special-purpose survey data collection is a rather expensive solution to
an information problem. Before launching such an effort, one should thoroughly
explore the potential for gathering the same information from existing records or from
other sources. Although some people think of a survey as a first effort to try to learn
20. CHAPTER 1 INTRODUCTION 3
something about a population, a full-scale probability sample survey should be
undertaken only after it is certain that the information cannot be obtained in other ways.
Even taking such a conservative approach, it is common to find that only a special-
purpose survey can provide the information that is needed. In addition to meeting needs
for data that are not available elsewhere, there are three potential properties of data
from a properly done survey that may make them preferable to data from other sources:
• Probability sampling enables one to have confidence that the sample is not a biased one and
to estimate how precise the data are likely to be. Data from a properly chosen sample are a
great improvement over data from a sample of those who attend meetings, speak loudest,
write letters, or happen to be convenient to poll.
• Standardized measurement that is consistent across all respondents ensures that comparable
information is obtained about everyone who is described. Without such measurement,
meaningful statistics cannot be produced.
• To meet analysis needs, a special-purpose survey may be the only way to ensure that all the
data needed for a given analysis are available and can be related. Even if there is informa-
tion about some set of events, it may not be paired with other characteristics needed to carry
out a desired analysis. For example, hospital discharge records invariably lack information
about income. Hence, a survey that collects both income and hospitalization data about
people is needed to study the relationship between a person’s income and hospitalization
experience.
There is always some information available on a given topic from what people say,
from impressions, or from records; also there are always imperfections in available
data. In addition to an assessment of information needs, the decision to do a survey also
should depend on available staff resources. Unless the needed staff and expertise, or the
resources to buy them, are available, the data resulting from a survey may not be very
good. That brings us to the topic of the next section: What constitutes a good survey?
COMPONENTS OF SURVEYS
Like all measures in all sciences, social survey measurement is not error free. The pro-
cedures used to conduct a survey have a major effect on the likelihood that the resulting
data will describe accurately what they are intended to describe.
A sample survey brings together three different methodologies: sampling, designing
questions, and data collection. Each of these activities has many applications outside of
sample surveys, but their combination is essential to good survey design.
Sampling
A census means gathering information about every individual in a population.
A major development in the process of making surveys useful was learning how to
21. 4 SURVEY RESEARCH METHODS
sample: to select a small subset of a population representative of the whole population.
The keys to good sampling are finding a way to give all (or nearly all) population mem-
bers the same (or a known) chance of being selected and using probability methods for
choosing the sample. Early surveys and polls often relied on samples of convenience or
on sampling from lists that excluded significant portions of the population. These did
not provide reliable, credible figures.
The U.S. Department of Agriculture actually developed the procedures for drawing the
comprehensive probability samples needed to provide statistically reliable descriptions of
populations living in a definable area. The procedures evolved from work designed to sam-
ple land areas for predicting crop yields; sampling housing units and the people living in
those housing units was simply an extension of that work. During World War II, a group of
social scientists was housed in the Department of Agriculture to do social surveys related to
the war effort. It was then that area probability sampling became firmly entrenched for
sampling general populations in social surveys.Area probability sampling is still the method
of choice for personal interview surveys of households. Converse (2009) provides an excel-
lent description of the evolution of survey methods in the United States.
Strategies for sampling have been refined since 1950. One major advance was the devel-
opment of random-digit dialing (RDD), which permitted the inclusion of households in
telephone surveys that did not have listed telephone numbers (Waksberg, 1978). The prin-
ciples of good sampling practice, however, have been well developed for a long time.
Question Design
Using questions as measures is another essential part of the survey process. The
initial survey efforts, representing extensions of journalism, were not careful about the
way that questions were posed. It soon became apparent, however, that sending an
interviewer out with a set of question objectives without providing specific wording for
the questions produced important differences in the answers that were obtained. Thus,
early in the 20th century, researchers began to write standardized questions for measur-
ing subjective phenomena. Again, researchers at the U.S. Department of Agriculture are
given credit for extending the use of standardized questions in the 1940s to situations
in which factual or objective information was sought. Payne (1951) published a land-
mark book providing practical guidelines for writing clear questions that interviewers
could administer as worded. Likert (1932) generally is credited for building a bridge
between the elaborate scaling techniques developed by psychophysical psychologists
for measuring subjective phenomena (e.g., Thurstone & Chave, 1929) and the practical
requirements of applied social survey research.
The major advance in question design in the last 20 years has been improved strate-
gies for evaluating questions. More than before, researchers now evaluate questions to
find out if they are well understood and if the answers are meaningful (see Presser
et al., 2004; Madans, Miller, Maitland, & Willis 2011). Pretests of surveys have become
more systematic, using analyses of tape-recorded interviews to identify problem ques-
tions. As a result, the choice of question wording is becoming more objective and less
a matter of research judgment.
22. CHAPTER 1 INTRODUCTION 5
Interviewing
Although not all surveys involve interviewing (as many surveys have respondents
answer self-administered questions in paper forms or on computers), it certainly is com-
mon to use an interviewer to ask questions and record answers. When interviewers are
used, it is important that they avoid influencing the answers respondents give, at the
same time maximizing the accuracy with which questions are answered.
The first major step in increasing interviewer consistency was to give them standard-
ized questions. It subsequently was found that interviewers also needed to be trained in
how to administer a survey to avoid introducing important biases in the answers they
obtained (Friedman, 1942). Hyman, Feldman, and Stember (1954) published a series of
studies documenting ways other than question wording that interviewers could influ-
ence the answers they obtained. Their work led to more elaborate training of interview-
ers with respect to strategies for probing when incomplete answers are obtained and for
handling the interpersonal aspects of the interview in nonbiasing ways. Cannell,
Oksenberg, and Converse (1977) advanced the process of trying to reduce between-
interviewer variation by specifically scripting the introductions and encouragement that
interviewers provide to respondents, while limiting unstructured discussion. The impor-
tance of interviewer training and supervision for ensuring data quality has been well
documented (Billiet & Loosveldt, 1988; Fowler & Mangione, 1990).
Mode of Data Collection
Until the 1970s, most academic and government surveys were done by in-person,
household interviewers. When telephone ownership became nearly universal in the
United States, telephone interviewing became a major mode of data collection. The
current frontier for data collection is the Internet. At the moment, its use is limited
because Internet access is still not universal in the United States and because the lists
and strategies for sampling e-mail addresses are limited. However, as access increases
and sampling strategies evolve, the use of the Internet to collect survey data is rapidly
increasing. Mail surveys, which in the past were used primarily when good address lists
were available for a target population, are also being used more widely as good quality
lists of addresses for the whole population have become available. Thus, more than
ever, researchers are making choices about the mode of data collection that will cost-
effectively produce the best quality data.
Total Survey Design
In many ways, the principles for good research practice were well developed in
the 1950s. However, understandably, the procedures and tools have changed in
response to new technologies and scientific advances. In some cases, we lack good
studies of how best to collect data for a particular purpose. However, even when
best practices have been well established, there is variability in the quality of the
procedures that are used.
23. 6 SURVEY RESEARCH METHODS
There are many reasons for variation in the quality of surveys. For some surveys,
imprecise figures will suffice. Lack of funding and of adequate staff, as well as lack of
methodological knowledge, no doubt all contribute to poor practice in some cases.
There also are some controversies about the value of strict probability sampling and
standardized question wording (see Converse, 2009; Groves, 2004; Schober & Conrad,
1997). One feature of survey research design that is partly to blame, however, is the
failure of researchers to put together high-quality procedures in all three of the salient
areas; it is not uncommon to see researchers attend carefully to some aspects of good
survey design while at the same time they neglect others. A critical orientation of this
book is the so-called total survey design perspective.
Every survey involves a number of decisions that have the potential to enhance or
detract from the accuracy (or precision) of survey estimates. Generally, the decisions
that would lead one to have better data involve more money, time, or other resources.
Thus the design of a survey involves a set of decisions to optimize the use of resources.
Optimal design will take into account all the salient aspects of the survey process.
With respect to sampling, critical issues include the following:
• the choice of whether or not to use a probability sample
• the sample frame (those people who actually have a chance to be sampled)
• the size of the sample
• the sample design (the particular strategy used for sampling people or households)
• the rate of response (the percentage of those sampled for whom data are actually collected)
With respect to question design, the researcher must decide the extent to which previ-
ous literature regarding the reliability and validity of questions will be drawn upon, the
use of consultants who are experts in question design, and the investment made in
pretesting and question evaluation. With respect to interviewers, researchers have
choices to make about the amount and kind of training and supervision to give. A design
decision cutting across all these areas is the mode of data collection: whether the
researcher will collect data by telephone, by mail, by personal interview, over the
Internet, or in some other way. The decision about which mode of data collection to use
has important cost implications and affects the quality of the data that will be collected.
These pieces, taken together, constitute what is called the total survey design. The
components of the design are interrelated in two important ways. First, the quality of
data will be no better than the most error-prone feature of the survey design. In the past,
researchers sometimes have focused on one or two features of the survey, such as the
size of the sample or the response rate, to evaluate the likely quality of data. Current
best practice, however, requires examination of all of the previously mentioned design
features. Indeed, Biemer (2010) has extended this thinking to include other aspects of
the resulting data, including timeliness and acceptability to potential user audiences. If
there is a major compromise or weakness in any aspect of the survey design, major
investments in other portions of the survey are not sensible. For example, if one is ask-
ing questions that respondents are unlikely to be able to answer with great precision, a
25. f
cut when the moon is wasting away would soon rot, whereas posts cut
while the moon is waxing are very durable.361
The same rule is observed for
the same reason in some parts of Germany.362
But the partisans of the ordinarily received opinion have sometimes
supported it by another reason, which introduces us to the second of those
fallacious appearances by which men have been led to regard the moon as
the cause of growth in plants. From observing rightly that dew falls most
thickly on cloudless nights, they inferred wrongly that it was caused by the
moon, a theory which the poet Alcman expressed in mythical form by saying
that dew was a daughter of Zeus and the moon.363
Hence the ancients
concluded that the moon is the great source of moisture, as the sun is the
great source of heat.364
And as the humid power of the moon was assumed
to be greater when the planet was waxing than when it was waning, they
thought that timber cut during the increase of the luminary would be
saturated with moisture, whereas timber cut in the wane would be
comparatively dry. Hence we are told that in antiquity carpenters would
reject timber felled when the moon was growing or full, because they
believed that such timber teemed with sap;365
and in the Vosges at the
present day people allege that wood cut at the new moon does not dry.366
We have seen that the same reason is assigned for the same practice in
Colombia.367
In the Hebrides peasants [pg 138] give the same reason for
cutting their peats when the moon is on the wane; “for they observe that if
they are cut in the increase, they continue still moist and never burn clear,
nor are they without smoke, but the contrary is daily observed of peats cut
in the decrease.”368
Thus misled by a double fallacy primitive philosophy comes to view the
moon as the great cause of vegetable growth, first, because the planet
seems itself to grow, and second, because it is supposed to be the source of
dew and moisture. It is no wonder, therefore, that agricultural peoples
should adore the planet which they believe to influence so profoundly the
crops on which they depend for subsistence. Accordingly we find that in the
hotter regions of America, where maize is cultivated and manioc is the
staple food, the moon was recognized as the principal object of worship,
and plantations of manioc were assigned to it as a return for the service it
rendered in the production of the crops. The worship of the moon in
preference to the sun was general among the Caribs, and, perhaps, also
among most of the other Indian tribes who cultivated maize in the tropical
26. forests to the east of the Andes; and the same thing has been observed,
under the same physical conditions, among the aborigines of the hottest
region of Peru, the northern valleys of Yuncapata. Here the Indians of
Pacasmayu and the neighbouring valleys revered the moon as their principal
divinity. The “house of the moon” at Pacasmayu was the chief temple of the
district; and the same sacrifices of maize-flour, of wine, and of children
which were offered by the mountaineers of the Andes to the Sun-god, were
offered by the lowlanders to the Moon-god in order that he might cause
their crops to thrive.369
In ancient [pg 139] Babylonia, where the population
was essentially agricultural, the moon-god took precedence of the sun-god
and was indeed reckoned his father.370
Hence it would be no matter for surprise if, after worshipping the crops
which furnished them with the means of subsistence, the ancient Egyptians
should in later times have identified the spirit of the corn with the moon,
which a false philosophy had taught them to regard as the ultimate cause of
the growth of vegetation. In this way we can understand why in their most
recent forms the myth and ritual of Osiris, the old god of trees and corn,
should bear many traces of efforts made to bring them into a superficial
conformity with the new doctrine of his lunar affinity.
[pg 140]
27. Chapter IX. The Doctrine of Lunar Sympathy.
In the preceding chapter some evidence was adduced of the sympathetic
influence which the waxing or waning moon is popularly supposed to exert
on growth, especially on the growth of vegetation. But the doctrine of lunar
sympathy does not stop there; it is applied also to the affairs of man, and
various customs and rules have been deduced from it which aim at the
amelioration and even the indefinite extension of human life. To illustrate
this application of the popular theory at length would be out of place here,
but a few cases may be mentioned by way of specimen.
The natural fact on which all the customs in question seem to rest is the
apparent monthly increase and decrease of the moon. From this observation
men have inferred that all things simultaneously wax or wane in sympathy
with it.371
Thus the Mentras or Mantras of the Malay Peninsula have a
tradition that in the beginning men did not die but grew thin with the
waning of the moon, and waxed fat as she neared the full.372
Of the Scottish
Highlanders we are told that “the moon in her increase, full growth, and in
her wane are with them the emblems of a rising, flourishing, and declining
fortune. At the last period of her revolution they carefully avoid to engage in
any business of importance; but the first and middle they seize with avidity,
presaging the most auspicious issue to their undertakings.”373
Similarly [pg
141] in some parts of Germany it is commonly believed that whatever is
undertaken when the moon is on the increase succeeds well, and that the
full moon brings everything to perfection; whereas business undertaken in
the wane of the moon is doomed to failure.374
This German belief has come
down, as we might have anticipated, from barbaric times; for Tacitus tells us
that the Germans considered the new or the full moon the most auspicious
time for business;375
and Caesar informs us that the Germans despaired of
victory if they joined battle before the new moon.376
The Spartans seem to
have been of the same opinion, for it was a rule with them never to march
out to war except when the moon was full. The rule prevented them from
sending troops in time to fight the Persians at Marathon,377
and but for
Athenian valour this paltry superstition might have turned the scale of battle
28. f
and decided the destiny of Greece, if not of Europe, for centuries. The
Athenians themselves paid dear for a similar scruple: an eclipse of the moon
cost them the loss of a gallant fleet and army before Syracuse, and
practically sealed the fate of Athens, for she never recovered from the
blow.378
So heavy is the sacrifice which superstition demands of its votaries.
In this respect the Greeks were on a level with the negroes of the Sudan,
among whom, if a march has been decided upon during the last quarter of
the moon, the departure is always deferred until the first day of the new
moon. No chief would dare to undertake an expedition and lead out his
warriors before the appearance of the crescent. Merchants and private
persons observe the same rule on their journeys.379
In like manner the
Mandingoes of Senegambia pay great attention to the changes of the moon,
and think it very unlucky to begin a journey or any other work of
consequence in the last quarter.380
It is especially the appearance of the new moon, with [pg 142] its promise
of growth and increase, which is greeted with ceremonies intended to renew
and invigorate, by means of sympathetic magic, the life of man. Observers,
ignorant of savage superstition, have commonly misinterpreted such
customs as worship or adoration paid to the moon. In point of fact the
ceremonies of new moon are probably in many cases rather magical than
religious. The Indians of the Ucayali River in Peru hail the appearance of the
new moon with great joy. They make long speeches to her, accompanied
with vehement gesticulations, imploring her protection and begging that she
will be so good as to invigorate their bodies.381
On the day when the new
moon first appeared, it was a custom with the Indians of San Juan
Capistrano, in California, to call together all the young men for the purpose
of its celebration. “Correr la luna!” shouted one of the old men, “Come, my
boys, the moon! the moon!” Immediately the young men began to run
about in a disorderly fashion as if they were distracted, while the old men
danced in a circle, saying, “As the moon dieth, and cometh to life again, so
we also having to die will again live.”382
An old traveller tells us that at the
appearance of every new moon the negroes of the Congo clapped their
hands and cried out, sometimes falling on their knees, “So may I renew my
life as thou art renewed.” But if the sky happened to be clouded, they did
nothing, alleging that the planet had lost its virtue.383
A somewhat similar
custom prevails among the Ovambo of South-Western Africa. On the first
moonlight night of the new moon, young and old, their bodies smeared with
white earth, perhaps in imitation of the planet's silvery light, dance to the
29. moon and address to it wishes which they feel sure will be granted.384
We
may conjecture that among these wishes is a prayer for a renewal of life.
When a Masai sees the new moon he throws a twig or stone at it with his
left hand, and says, “Give me [pg 143] long life,” or “Give me strength”; and
when a pregnant woman sees the new moon she milks some milk into a
small gourd, which she covers with green grass. Then she pours the milk
away in the direction of the moon and says, “Moon, give me my child
safely.”385
Among the Wagogo of German East Africa, at sight of the new
moon some people break a stick in pieces, spit on the pieces, and throw
them towards the moon, saying, “Let all illness go to the west, where the
sun sets.”386
Among the Boloki of the Upper Congo there is much shouting
and gesticulation on the appearance of a new moon. Those who have
enjoyed good health pray that it may be continued, and those who have
been sick ascribe their illness to the coming of the luminary and beg her to
take away bad health and give them good health instead.387
The Esthonians
think that all the misfortunes which might befall a man in the course of a
month may be forestalled and shifted to the moon, if a man will only say to
the new moon, “Good morrow, new moon. I must grow young, you must
grow old. My eyes must grow bright, yours must grow dark. I must grow
light as a bird, you must grow heavy as iron.”388
On the fifteenth day of the
moon, that is, at the time when the luminary has begun to wane, the
Coreans take round pieces of paper, either red or white, which represent the
moon, and having fixed them perpendicularly on split sticks they place them
on the tops of the houses. Then persons who have been forewarned by
fortune-tellers of impending evil pray to the moon to remove it from
them.389
[pg 144]
In India people attempt to absorb the vital influence of the moon by
drinking water in which the luminary is reflected. Thus the Mohammedans
of Oude fill a silver basin with water and hold it so that the orb of the full
moon is mirrored in it. The person to be benefited must look steadfastly at
the moon in the basin, then shut his eyes and drink the water at one gulp.
Doctors recommend the draught as a remedy for nervous disorders and
palpitation of the heart. Somewhat similar customs prevail among the
Hindoos of Northern India. At the full moon of the month of Kuar
(September-October) people lay out food on the house-tops, and when it
has absorbed the rays of the moon they distribute it among their relations,
who are supposed to lengthen their life by eating of the food which has thus
30. f
f
f
been saturated with moonshine. Patients are often made to look at the
moon reflected in melted butter, oil, or milk as a cure for leprosy and the
like diseases.390
Naturally enough the genial influence of moonshine is often supposed to be
particularly beneficial to children; for will not the waxing moon help them to
wax in strength and stature? Hence in the island of Kiriwina, one of the
Trobriands Group to the east of New Guinea, a mother always lifts up or
presents her child to the first full moon after its birth in order that it may
grow fast and talk soon.391
So among the Baganda of Central Africa it was
customary for each mother to take her child out at the first new moon after
its birth, and to point out the moon to the infant; this was thought to make
the child grow healthy and strong.392
Among the Thonga of South Africa the
presentation of the baby to the moon does not take place until the mother
has resumed her monthly periods, which usually happens in the third month
after the birth. When the new moon appears, the mother takes a torch or a
burning brand from the fire and goes to the ash-heap behind the hut. She is
followed by the grandmother carrying the child. At the ash-heap the mother
throws the burning stick towards the moon, while the grandmother tosses
the [pg 145] baby into the air, saying, “This is your moon!” The child squalls
and rolls over on the ash-heap. Then the mother snatches up the infant and
nurses it; so they go home.393
The Guarayos Indians, who inhabit the gloomy tropical forests of Eastern
Bolivia, lift up their children in the air at new moon in order that they may
grow.394
Among the Apinagos Indians, on the Tocantins River in Brazil, the
French traveller Castelnau witnessed a remarkable dance by moonlight. The
Indians danced in two long ranks which faced each other, the women on
one side, the men on the other. Between the two ranks of dancers blazed a
great fire. The men were painted in brilliant colours, and for the most part
wore white or red skull-caps made of maize-flour and resin. Their dancing
was very monotonous and consisted of a jerky movement of the body, while
the dancer advanced first one leg and then the other. This dance they
accompanied with a melancholy song, striking the ground with their
weapons. Opposite them the women, naked and unpainted, stood in a
single rank, their bodies bent slightly forward, their knees pressed together,
their arms swinging in measured time, now forward, now backward, so as to
join hands. A remarkable figure in the dance was a personage painted
scarlet all over, who held in his hand a rattle composed of a gourd full of
pebbles. From time to time he leaped across the great fire which burned
31. between the men and the women. Then he would run rapidly in front of the
women, stopping now and then before one or other and performing a series
of strange gambols, while he shook his rattle violently. Sometimes he would
sink with one knee to the ground, and then suddenly throw himself
backward. Altogether the agility and endurance which he displayed were
remarkable. This dance lasted for hours. When a woman was tired out she
withdrew, and her place was taken by another; but the same men danced
the monotonous dance all night. Towards midnight the moon attained the
zenith and flooded the scene with her bright rays. A change [pg 146] now
took place in the dance. A long line of men and women advanced to the fire
between the ranks of the dancers. Each of them held one end of a
hammock in which lay a new-born infant, whose squalls could be heard.
These babes were now to be presented by their parents to the moon. On
reaching the end of the line each couple swung the hammock,
accompanying the movement by a chant, which all the Indians sang in
chorus. The song seemed to consist of three words, repeated over and over
again. Soon a shrill voice was heard, and a hideous old hag, like a skeleton,
appeared with her arms raised above her head. She went round and round
the assembly several times, then disappeared in silence. While she was
present, the scarlet dancer with the rattle bounded about more furiously
than ever, stopping only for a moment while he passed in front of the line of
women. His body was contracted and bent towards them, and described an
undulatory movement like that of a worm writhing. He shook his rattle
violently, as if he would fain kindle in the women the fire which burned in
himself. Then rising abruptly he would resume his wild career. During this
time the loud voice of an orator was heard from the village repeating a
curious name without cessation. Then the speaker approached slowly,
carrying on his back some gorgeous bunches of brilliant feathers and under
his arm a stone axe. Behind him walked a young woman bearing an infant
in a loose girdle at her waist; the child was wrapped in a mat, which
protected it against the chill night air. The couple paced slowly for a minute
or two, and then vanished without speaking a word. At the same moment
the curious name which the orator had shouted was taken up by the whole
assembly and repeated by them again and again. This scene in its turn
lasted a long time, but ceased suddenly with the setting of the moon. The
French traveller who witnessed it fell asleep, and when he awoke all was
calm once more: there was nothing to recall the infernal dances of the
night.395
32. f
In explanation of these dances Castelnau merely observes [pg 147] that the
Apinagos, like many other South American Indians, pay a superstitious
respect to the moon. We may suppose that the ceremonious presentation of
the infants to the moon was intended to ensure their life and growth. The
names solemnly chanted by the whole assembly were probably those which
the parents publicly bestowed on their children. As to the scarlet dancer
who leaped across the fire, we may conjecture that he personated the
moon, and that his strange antics in front of the women were designed to
impart to them the fertilizing virtue of the luminary, and perhaps to facilitate
their delivery.
Among the Baganda of Central Africa there is general rejoicing when the
new moon appears, and no work is done for seven days. When the crescent
is first visible at evening, mothers take out their babies and hold them at
arms' length, saying, “I want my child to keep in health till the moon
wanes.” At the same time a ceremony is performed which may be intended
to ensure the king's life and health throughout the ensuing month. It is a
custom with the Baganda to preserve the king's navel-string with great care
during his life. The precious object is called the “Twin” of the king, as if it
were his double; and the ghost of the royal afterbirth is believed to be
attached to it. Enclosed in a pot, which is wrapt in bark cloths, the navel-
string is kept in a temple specially built for it near the king's enclosure, and
a great minister of state acts as its guardian and priest. Every new moon, at
evening, he carries it in state, wrapped in bark cloths, to the king, who
takes it into his hands, examines it, and returns it to the minister. The
keeper of the navel-string then goes back with it to the house and sets it in
the doorway, where it remains all night. Next morning it is taken from its
wrappings and again placed in the doorway until the evening, when it is
once more swathed in bark cloths and restored to its usual place.396
Apparently the navel-string is conceived as a vital portion, a sort of external
soul, of the [pg 148] king; and the attentions bestowed on it at the new
moon may be supposed to refresh and invigorate it, thereby refreshing and
invigorating the king's life.
The Armenians appear to think that the moon exercises a baleful influence
on little children. To avert that influence a mother will show the moon to her
child and say, “Thine uncle, thine uncle.” For the same purpose the father
and mother will mount to the roof of the house at new moon on a
Wednesday or Friday. The father then puts the child on a shovel and gives it
to the mother, saying, “If it is thine, take it to thee. But if it is mine, rear it
33. f
and give it to me back.” The mother then takes the child and the shovel,
and returns them to the father in like manner.397
A similar opinion as to the
noxious influence of moonshine on children was apparently held by the
ancient Greeks; for Greek nurses took great care never to show their infants
to the moon.398
Some Brazilian Indians in like manner guard babies against
the moon, believing that it would make them ill. Immediately after delivery
mothers will hide themselves and their infants in the thickest parts of the
forest in order that the moonlight may not fall on them.399
It would be easy
to understand why the waning moon should be deemed injurious to
children; they might be supposed to peak and pine with its dwindling light.
Thus in Angus it is thought that if a child be weaned during the waning of
the moon, it will decay all the time that the moon continues to wane.400
But
it is less easy to see why the same deleterious influence on children should
be ascribed to moonlight in general.
There are many other ways in which people have sought to turn lunar
sympathy to practical account. Clearly the increase of the moon is the time
to increase your goods, and the decrease of the moon is the time to
diminish your ills. Acting on this imaginary law of nature many persons in
Europe show their money to the new moon or turn it in [pg 149] their
pockets at that season, in the belief that the money will grow with the
growth of the planet; sometimes, by way of additional precaution, they spit
on the coin at the same time.401
“Both Christians and Moslems in Syria turn
their silver money in their pockets at the new moon for luck; and two
persons meeting under the new moon will each take out a silver coin and
embrace, saying, ‘May you begin and end; and may it be a good month to
us.’ ”402
Conversely the waning of the moon is the most natural time to get
rid of bodily ailments. In Brittany they think that warts vary with the phases
of the moon, growing as it waxes and vanishing away as it wanes.403
Accordingly, they say in Germany that if you would rid yourself of warts you
should treat them when the moon is on the decrease.404
And a German cure
for toothache, earache, headache, and so forth, is to look towards the
waning moon and say, “As the moon decreases, so may my pains decrease
also.”405
However, some Germans reverse the rule. They say, for example,
that if you are afflicted with a wen, you should face the waxing moon, lay
your finger on the wen, and say thrice, “What I see waxes; what I touch, let
it vanish away.” After each of these two sentences you should cross yourself
thrice. Then go home without speaking to any one, and repeat three
paternosters behind the kitchen door.406
The Huzuls of the Carpathians
34. recommend a somewhat similar, and no doubt equally efficacious, cure for
waterbrash. They say that at new moon the patient should run thrice round
the house and then say to the moon, “Moon, moon, where wast thou?”
“Behind the mountain.” “What hast thou eaten there?” “Horse flesh.” “Why
hast thou brought me nothing?” “Because I forgot.” “May the waterbrash
[pg 150] forget to burn me!”407
Thus a curative virtue appears to be
attributed by some people to the waning and by others to the waxing moon.
There is perhaps just as much, or as little, to be said for the one attribution
as for the other.
[pg 151]
35. f
Chapter X. The King As Osiris.
In the foregoing discussion we found reason to believe that the Semitic
Adonis and the Phrygian Attis were at one time personated in the flesh by
kings, princes, or priests who played the part of the god for a time and then
either died a violent death in the divine character or had to redeem their life
in one way or another, whether by performing a make-believe sacrifice at
some expense of pain and danger to themselves, or by delegating the duty
to a substitute.408
Further, we conjectured that in Egypt the part of Osiris
may have been played by the king himself.409
It remains to adduce some
positive evidence of this personation.
A great festival called the Sed was celebrated by the Egyptians with much
solemnity at intervals of thirty years. Various portions of the ritual are
represented on the ancient monuments of Hieraconpolis and Abydos and in
the oldest decorated temple of Egypt known to us, that of Usirniri at Busiris,
which dates from the fifth dynasty. It appears that the ceremonies were as
old as the Egyptian civilization, and that they continued to be observed till
the end of the Roman period.410
The reason for holding them at intervals of
thirty [pg 152] years is uncertain, but we can hardly doubt that the period
was determined by astronomical considerations. According to one view, it
was based on the observation of Saturn's period of revolution round the
sun, which is, roughly speaking, thirty years, or, more exactly, twenty-nine
years and one hundred and seventy-four days.411
According to another view,
the thirty years' period had reference to Sirius, the star of Isis. We have
seen that on account of the vague character of the old Egyptian year the
heliacal rising of Sirius shifted its place gradually through every month of
the calendar.412
In one hundred and twenty years the star thus passed
through one whole month of thirty days. To speak more precisely, it rose on
the first of the month during the first four years of the period: it rose on the
second of the month in the second four years, on the third of the month in
the third four years; and so on successively, till in the last four years of the
36. hundred and twenty years it rose on the last day of the month. As the
Egyptians watched the annual summer rising of the star with attention and
associated it with the most popular of their goddesses, it would be natural
that its passage from one month to another, at intervals of one hundred and
twenty years, should be the occasion of a great festival, and that the long
period of one hundred and twenty years should be divided into four minor
periods of thirty years respectively, each celebrated by a minor festival.413
If
this theory of the Sed festivals is correct, we should expect to find that
every fourth celebration was distinguished from the rest by a higher degree
of solemnity, since it marked the completion of a twelfth part of the star's
journey through the twelve [pg 153] months. Now it appears that in point of
fact every fourth Sed festival was marked off from its fellows by the
adjective tep or “chief,” and that these “chief” celebrations fell as a rule in
the years when Sirius rose on the first of the month.414
These facts confirm
the view that the Sed festival was closely connected with the star Sirius, and
through it with Isis.
However, we are here concerned rather with the meaning and the rites of
the festival than with the reasons for holding it once every thirty years. The
intention of the festival seems to have been to procure for the king a new
lease of life, a renovation of his divine energies, a rejuvenescence. In the
inscriptions of Abydos we read, after an account of the rites, the following
address to the king: “Thou dost recommence thy renewal, thou art granted
to flourish again like the infant god Moon, thou dost grow young again, and
that from season to season, like Nun at the beginning of time, thou art born
again in renewing the Sed festivals. All life comes to thy nostril, and thou art
king of the whole earth for ever.”415
In short, on these occasions it appears
to have been supposed that the king was in a manner born again.
But how was the new birth effected? Apparently the essence of the rites
consisted in identifying the king with Osiris; for just as Osiris had died and
risen again from the dead, so the king might be thought to die and to live
again with the god whom he personated. The ceremony would thus be for
the king a death as well as a rebirth. Accordingly in pictures of the Sed
festival on the monuments we see the king posing as the dead Osiris. He
sits in a shrine like a god, holding in his hands the crook and flail of Osiris:
he is wrapped in tight bandages like the mummified Osiris; indeed, there is
nothing but his name to prove that he is not Osiris himself. This
enthronement of the king in the attitude of the dead god seems to have
been the principal event of the festival.416
Further, the queen and the king's
37. daughters figured prominently in the ceremonies.417
A [pg 154] discharge of
arrows formed part of the rites;418
and in some sculptures at Carnac the
queen is portrayed shooting arrows towards the four quarters of the world,
while the king does the same with rings.419
The oldest illustration of the
festival is on the mace of Narmer, which is believed to date from 5500 b.c.
Here we see the king seated as Osiris in a shrine at the top of nine steps.
Beside the shrine stand fan-bearers, and in front of it is a figure in a
palanquin, which, according to an inscription in another representation of
the scene, appears to be the royal child. An enclosure of curtains hung on
poles surrounds the dancing-ground, where three men are performing a
sacred dance. A procession of standards is depicted beside the enclosure; it
is headed by the standard of the jackal-god Up-uat, the “opener of ways”
for the dead.420
Similarly on a seal of King Zer, or rather Khent, one of the
early kings of the first dynasty, the monarch appears as Osiris with the
standard of the jackal-god before him. In front of him, too, is the ostrich
feather on which “the dead king was supposed to ascend into heaven. Here,
then, the king, identified with Osiris, king of the dead, has before him the
jackal-god, who leads the dead, and the ostrich feather, which symbolizes
his reception into the sky.”421
There are even grounds for thinking that in
order to complete the mimic death of the king at the Sed festival an effigy
of him, clad in the costume of Osiris, was solemnly buried in a cenotaph.422
According to Professor Flinders Petrie, “the conclusion may be drawn thus.
In the savage age of prehistoric times, the Egyptians, like many other
African and Indian peoples, killed their priest-king at stated intervals, in
order that the ruler should, with unimpaired life and health, be enabled to
maintain the kingdom in its highest condition. The royal daughters were
present in order that they might be married to his successor. The jackal-god
went before [pg 155] him, to open the way to the unseen world; and the
ostrich feather received and bore away the king's soul in the breeze that
blew it out of sight. This was the celebration of the ‘end,’ the sed feast. The
king thus became the dead king, patron of all those who had died in his
reign, who were his subjects here and hereafter. He was thus one with
Osiris, the king of the dead. This fierce custom became changed, as in other
lands, by appointing a deputy king to die in his stead; which idea survived in
the Coptic Abu Nerūs, with his tall crown of Upper Egypt, false beard, and
sceptre. After the death of the deputy, the real king renewed his life and
reign. Henceforward this became the greatest of the royal festivals, the
38. apotheosis of the king during his life, after which he became Osiris upon
earth and the patron of the dead in the underworld.”423
A similar theory of the Sed festival is maintained by another eminent
Egyptologist, M. Alexandre Moret. He says: “In most of the temples of
Egypt, of all periods, pictures set forth for us the principal scenes of a
solemn festival called ‘festival of the tail,’ the Sed festival. It consisted
essentially in a representation of the ritual death of the king followed by his
rebirth. In this case the king is identified with Osiris, the god who in
historical times is the hero of the sacred drama of humanity, he who guides
us through the three stages of life, death, and rebirth in the other world.
Hence, clad in the funeral costume of Osiris, with the tight-fitting garment
clinging to him like a shroud, Pharaoh is conducted to the tomb; and from it
he returns rejuvenated and reborn like Osiris emerging from the dead. How
was this fiction carried out? how was this miracle performed? By the
sacrifice of human or animal victims. On behalf of the king a priest lay down
in the skin of the animal victim: he assumed the posture characteristic of an
embryo in its mother's womb: when he came forth from the skin he was
deemed to be reborn; and Pharaoh, for whom this rite was celebrated, was
himself reborn, or to adopt the Egyptian expression, [pg 156] ‘he renewed
his births.’ And in testimony of the due performance of the rites the king girt
his loins with the tail, a compendious representative of the skin of the
sacrificed beast, whence the name of ‘the festival of the tail.’
“How are we to explain the rule that at a certain point of his reign every
Pharaoh must undergo this ritual death followed by fictitious rebirth? Is it
simply a renewal of the initiation into the Osirian mysteries? or does the
festival present some more special features? The ill-defined part played by
the royal children in these rites seems to me to indicate that the Sed festival
represents other episodes which refer to the transmission of the regal office.
At the dawn of civilization in Egypt the people were perhaps familiar with
the alternative either of putting their king to death in his full vigour in order
that his power should be transmitted intact to his successor, or of
attempting to rejuvenate him and to ‘renew his life.’ The latter measure was
an invention of the Pharaohs. How could it be carried out more effectively
than by identifying themselves with Osiris, by applying to themselves the
process of resurrection, the funeral rites by which Isis, according to the
priests, had magically saved her husband from death? Perhaps the fictitious
death of the king may be regarded as a mitigation of the primitive murder of
the divine king, a transition from a barbarous reality to symbolism.”424
39. f
[pg 157]
Whether this interpretation of the Sed festival be accepted in all its details
or not, one thing seems quite certain: on these solemn occasions the god
Osiris was personated by the king of Egypt himself. That is the point with
which we are here chiefly concerned.
[pg 158]
40. f
Chapter XI. The Origin of Osiris.
Thus far we have discussed the character of Osiris as he is presented to us
in the art and literature of Egypt and in the testimonies of Greek writers;
and we have found that judged by these indications he was in the main a
god of vegetation and of the dead. But we have still to ask, how did the
conception of such a composite deity originate? Did it arise simply through
observation of the great annual fluctuations of the seasons and a desire to
explain them? Was it a result of brooding over the mystery of external
nature? Was it the attempt of a rude philosophy to lift the veil and explore
the hidden springs that set the vast machine in motion? That man at a very
early stage of his long history meditated on these things and evolved certain
crude theories which partially satisfied his craving after knowledge is
certain; from such meditations of Babylonian and Phrygian sages appear to
have sprung the pathetic figures of Adonis and Attis; and from such
meditations of Egyptian sages may have sprung the tragic figure of Osiris.
Yet a broad distinction seems to sever the myth and worship of Osiris from
the kindred myths and worships of Adonis and Attis. For while Adonis and
Attis were minor divinities in the religion of Western Asia, completely
overshadowed by the greater deities of their respective pantheons, the
solemn figure of Osiris towered in solitary grandeur over all the welter of
Egyptian gods, like a pyramid of his native land lit up by the last rays of the
setting sun when all below it is in shadow. And whereas legend generally
represented Adonis and Attis as simple swains, mere herdsmen [pg 159] or
hunters whom the fatal love of a goddess had elevated above their homely
sphere into a brief and melancholy pre-eminence, Osiris uniformly appears
in tradition as a great and beneficent king. In life, he ruled over his people,
beloved and revered for the benefits he conferred on them and on the
world; in death he reigned in their hearts and memories as lord of the dead,
the awful judge at whose bar every man must one day stand to give an
account of the deeds done in the body and to receive the final award. In the
faith of the Egyptians the cruel death and blessed resurrection of Osiris
occupied the same place as the death and resurrection of Christ hold in the
41. f faith of Christians. As Osiris died and rose again from the dead, so they
hoped through him and in his dear name to wake triumphant from the sleep
of death to a blissful eternity. That was their sheet-anchor in life's stormy
sea; that was the hope which supported and consoled millions of Egyptian
men and women for a period of time far longer than that during which
Christianity has now existed on earth. In the long history of religion no two
divine figures resemble each other more closely in the fervour of personal
devotion which they have kindled and in the high hopes which they have
inspired than Osiris and Christ. The sad figure of Buddha indeed has been
as deeply loved and revered by countless millions; but he had no glad
tidings of immortality for men, nothing but the promise of a final release
from the burden of mortality.
And if Osiris and Christ have been the centres of the like enthusiastic
devotion, may not the secret of their influence have been similar? If Christ
lived the life and died the death of a man on earth, may not Osiris have
done so likewise? The immense and enduring popularity of his worship
speaks in favour of the supposition; for all the other great religious or semi-
religious systems which have won for themselves a permanent place in the
affections of mankind, have been founded by individual great men, who by
their personal life and example exerted a power of attraction such as no
cold abstractions, no pale products of the collective wisdom or folly could
ever exert on the minds and hearts of humanity. Thus it was with Buddhism,
with [pg 160] Confucianism, with Christianity, and with Mohammedanism;
and thus it may well have been with the religion of Osiris. Certainly we shall
do less violence to the evidence if we accept the unanimous tradition of
ancient Egypt on this point than if we resolve the figure of Osiris into a myth
pure and simple. And when we consider that from the earliest to the latest
times Egyptian kings were worshipped as gods both in life and in death,
there appears to be nothing extravagant or improbable in the view that one
of them by his personal qualities excited a larger measure of devotion than
usual during his life and was remembered with fonder affection and deeper
reverence after his death; till in time his beloved memory, dimmed,
transfigured, and encircled with a halo of glory by the mists of time, grew
into the dominant religion of his people. At least this theory is reasonable
enough to deserve a serious consideration. If we accept it, we may suppose
that the mythical elements, which legend undoubtedly ascribed to Osiris,
were later accretions which gathered about his memory like ivy about a ruin.
There is no improbability in such a supposition; on the contrary, all analogy
42. f
f
f
is in its favour, for nothing is more certain than that myths grow like weeds
round the great historical figures of the past.
In recent years the historical reality of Osiris as a king who once lived and
reigned in Egypt has been maintained by more than one learned scholar;425
and without venturing to pronounce a decided opinion on so obscure and
difficult a question, I think it worth while, following the example of Dr. Wallis
Budge, to indicate certain modern African analogies which tend to confirm
the view that beneath the mythical wrappings of Osiris there lay the
mummy of a dead man. At all events the analogies which I shall cite suffice
to prove that the custom of worshipping dead kings has not been confined
to Egypt, but has been apparently widespread throughout Africa, though the
evidence now at our disposal only enables us to detect the observance of
the [pg 161] custom at a few points of the great continent. But even if the
resemblance in this respect between ancient Egypt and modern Africa
should be regarded as established, it would not justify us in inferring an
ethnical affinity between the fair or ruddy Egyptians and the black aboriginal
races who occupy almost the whole of Africa except a comparatively narrow
fringe on the northern sea-board. Scholars are still divided on the question
of the original home and racial relationship of the ancient Egyptians. It has
been held on the one hand that they belong to an indigenous white race
which has been always in possession of the Mediterranean coasts of Africa;
and on the other hand it has been supposed that they are akin to the
Semites in blood as well as in language, and that they entered Africa from
the East, whether by gradual infiltration or on a sudden wave of conquest
like the Arabs in the decline of the Roman empire.426
On either view a great
gulf divided them from the swarthy natives of the Sudan, with whom they
were always in contact on their southern border; and though a certain
admixture may have taken place through marriage between the two races, it
seems unsafe to assume that the religious and political resemblances which
can be traced between them are based on any closer relationship than the
general similarity in structure and functions of the human mind.
In a former part of this work we saw that the Shilluks, a pastoral and
partially agricultural people of the White Nile, worship the spirits of their
dead kings.427
The graves of the deceased monarchs form indeed the
national or tribal [pg 162] temples; and as each king is interred at the
village where he was born and where his afterbirth is buried, these grave-
shrines are scattered over the country. Each of them usually comprises a
small group of round huts, resembling the common houses of the people,
43. the whole being enclosed by a fence; one of the huts is built over the grave,
the others are occupied by the guardians of the shrine, who at first are
generally the widows or old men-servants of the deceased king. When these
women or retainers die, they are succeeded in office by their descendants,
for the tombs are maintained in perpetuity, so that the number of temples
and of gods is always on the increase. Cattle are dedicated to these royal
shrines and animals sacrificed at them. For example, when the millet crop
threatens to fail or a murrain breaks out among the beasts, one of the dead
kings will appear to somebody in a dream and demand a sacrifice. The
dream is reported to the king, and he immediately orders a bullock and a
cow to be sent to the grave of the dead king who appeared in a vision of
the night to the sleeper. This is done; the bullock is killed and the cow
added to the sacred herd of the shrine. It is customary, also, though not
necessary, at harvest to offer some of the new millet at the temple-tombs of
the kings; and sick people send animals to be sacrificed there on their
behalf. Special regard is paid to trees that grow near the graves of the
kings; and the spirits of the departed monarchs are believed to appear from
time to time in the form of certain animals. One of them, for example,
always takes the shape of a certain insect, which seems to be the larva of
the Mantidae. When a Shilluk finds one of these insects, he will take it up in
his hands and deposit it reverentially at the shrine. Other kings manifest
themselves as a certain species of white birds; others assume the form of
giraffes. When one of these long-legged and long-necked creatures comes
stalking up fearlessly to a village where there is a king's grave, the people
know that the king's soul is in the animal, and the attendants at the royal
tomb testify their joy at the appearance of their master by sacrificing a
sheep or even a bullock.
But of all the dead kings none is revered so deeply or [pg 163] occupies so
large a place in the minds of the people as Nyakang, the traditional founder
of the dynasty and the ancestor of all the kings who have reigned after him
to the present day. Of these kings the Shilluks have preserved the memory
and the genealogy; twenty-six seem to have sat on the throne since
Nyakang, but the period of time covered by their reigns is much shorter
than it would have been under conditions such as now prevail in Europe; for
down to the time when their country came under British rule it was the
regular custom of the Shilluks to put their kings to death as soon as they
showed serious symptoms of bodily or mental decay. The custom was based
on “the conviction that the king must not be allowed to become ill or senile,
lest with his diminishing vigour the cattle should sicken and fail to bear their
44. increase, the crops should rot in the fields, and man, stricken with disease,
should die in ever-increasing numbers.”428
It is said that Nyakang, like
Romulus, disappeared in a great storm, which scattered all the people about
him; in their absence the king took a cloth, tied it tightly round his neck,
and strangled himself. According to one account, that is the death which all
his successors on the throne have died;429
but while tradition appears to be
unanimous as to the custom of regicide, it varies as to the precise mode in
which the kings were relieved of their office and of life. But still the people
are convinced that Nyakang did not really die but only vanished mysteriously
away like the wind. When a missionary asked the Shilluks as to the manner
of Nyakang's death, they were filled with amazement at his ignorance and
stoutly maintained that he never died, for were he to die all the Shilluks
would die also.430
The graves of this deified king are shown in various parts
of the country.
From time to time the spirit of Nyakang manifests itself to his people in the
form of an animal. Any creature of regal port or surpassing beauty may
serve as his temporary incarnation. Such among wild animals are lions,
crocodiles, little yellow snakes that crawl about men's houses, the finest
sorts of antelopes, flamingoes with their rose-pink and scarlet [pg 164]
plumage, and butterflies of all sorts with their brilliant and varied hues. An
unusually fine head of cattle is also recognized as the abode of the great
king's soul; for example he once appeared in the shape of a white bull,
whereupon the living king commanded special sacrifices to be offered in
honour of his deified predecessor. When a bird in which the royal spirit is
known to be lodged lights on a tree, that tree becomes sacred to Nyakang;
beads and cloths are hung on its boughs, sacrifices and prayers are offered
below it. Once when the Turks unknowingly felled such a tree, fear and
horror fell on the Shilluks who beheld the sacrilege. They filled the air with
lamentations and killed an ox to appease their insulted ancestor.431
Particular
regard is also paid to trees that grow near the graves of Nyakang, though
they are not regularly worshipped.432
In one place two gigantic baobab trees
are pointed out as marking the spot where Nyakang once stood, and
sacrifices are now offered under their spreading shade.433
There seems to be no doubt that in spite of the mythical elements which
have gathered round his memory, Nyakang was a real man, who led the
Shilluks to their present home on the Nile either from the west or from the
south; for on this point tradition varies. “The first and most important
ancestor, who is everywhere revered, is Nyakang, the first Shilluk king. He
45. always receives the honourable titles of Father (uò), Ancestor (qua), King
(red) or Kings (ror), Ancestors, and Great Man Above (čal duong mal) to
distinguish him from the other great men on earth. Nyakang, as we know,
was an historical personage; he led the Shilluks to the land which they now
occupy; he helped them to victory, made them great and warlike, regulated
marriage and law, distributed the country among them, divided it into
districts, and in order to increase the dependence of the people on him and
to show them his power, became their greatest benefactor by giving himself
out as the bestower of rain.”434
Yet Nyakang is now universally revered by
the people as a demi-god; indeed for all practical purposes [pg 165] his
worship quite eclipses that of the supreme god Juok, the creator, who,
having ordered the world, committed it to the care of ancestral spirits and
demons, and now, dwelling aloft, concerns himself no further with human
affairs. Hence men pay little heed to their creator and seldom take his name
into their lips except in a few conventional forms of salutation at meeting
and parting like our “Good-bye.” Far otherwise is it with Nyakang. He “is the
ancestor of the Shilluk nation and the founder of the Shilluk dynasty. He is
worshipped, sacrifices and prayers are offered to him; he may be said to be
lifted to the rank of a demi-god, though they never forget that he has been
a real man. He is expressly designated as ‘little’ in comparison with God.”
Yet “in the political, religious and personal life Nyakang takes a far more
important place than Juok. Nyakang is the national hero, of whom each
Shilluk feels proud, who is praised in innumerable popular songs and
sayings; he is not only a superior being, but also a man. He is the sublime
model for every true Shilluk; everything they value most in their national
and private life has its origin in him: their kingdom and their fighting as well
as cattle-breeding and farming. While Nyakang is their good father, who
only does them good, Juok is the great, uncontrollable power, which is to be
propitiated, in order to avoid his inflictions of evil.”435
Indeed “the whole
working religion of the Shilluk is a cult of Nyakang, the semi-divine ancestor
of their kings, in each of whom his spirit is immanent.”436
The transmission
of the divine or semi-divine spirit of Nyakang to the reigning monarch
appears to take place at the king's installation and to be effected by means
of a rude wooden effigy of Nyakang, in which the spirit of that deified man
is perhaps supposed to be immanent. But however the spiritual transmission
may be carried out, “the fundamental idea of the cult of the Shilluk divine
kings is the immanence in each of the spirit of Nyakang.”437
Thus the Shilluk
kings are encircled with a [pg 166] certain halo of divinity because they are
46. f
f
f
thought to be animated by the divine spirit of their ancestor, the founder of
the dynasty.
The universal belief of the Shilluks in the former humanity of Nyakang is
strongly confirmed by the exact parallelism which prevails between his
worship and that of the dead kings his successors. Like them he is
worshipped at his tomb; but unlike them he has not one tomb only, but ten
scattered over the country. Each of these tombs is called “the grave of
Nyakang,” though the people well know that nobody is buried there. Like
the grave-shrines of the other kings, those of Nyakang consist of a small
group of circular huts of the ordinary pattern enclosed by a fence. Only
children under puberty and the few old people whose duty it is to take care
of the shrines may enter these sacred enclosures. The rites performed at
them resemble those observed at the shrines of the kings. Two great
ceremonies are annually performed at the shrines of Nyakang: one is
observed before the beginning of the rainy season in order to ensure a due
supply of rain; the other is a thanksgiving at harvest, when porridge made
from the new grain is poured out on the threshold of Nyakang's hut and
smeared on the outer walls of the building. Even before the millet is reaped
the people cut some of the ripening ears and thrust them into the thatch of
the sacred hut. Thus it would seem that the Shilluks believe themselves to
be dependent on the favour of Nyakang for the rain and the crops. “As the
giver of rain, Nyakang is the first and greatest benefactor of the people. In
that country rain is everything, without rain there is nothing. The Shilluk
does not trouble his head about artificial irrigation, he waits for the rain. If
the rain falls, then the millet grows, the cows thrive, man has food and can
dance and marry; for that is the ideal of the Shilluks.”438
Sick people also
bring or send sheep as an offering to the nearest shrine of Nyakang in order
that they may be healed of their sickness. The attendants of the [pg 167]
sanctuary slaughter the animal, consume its flesh, and give the sufferer the
benefit of their prayers.439
The example of Nyakang seems to show that under favourable
circumstances the worship of a dead king may develop into the dominant
religion of a people. There is, therefore, no intrinsic improbability in the view
that in ancient Egypt the religion of Osiris originated in that way. Certainly
some curious resemblances can be traced between the dead Nyakang and
the dead Osiris. Both died violent and mysterious deaths: the graves of both
were pointed out in many parts of the country: both were deemed the great
sources of fertility for the whole land: and both were associated with certain
47. f
f
sacred trees and animals, particularly with bulls. And just as Egyptian kings
identified themselves both in life and in death with their deified predecessor
Osiris, so Shilluk kings are still believed to be animated by the spirit of their
deified predecessor Nyakang and to share his divinity.
Another African people who regularly worship, or rather used to worship,
the spirits of their dead kings are the Baganda. Their country Uganda lies at
the very source of the Nile, where the great river issues from Lake Victoria
Nyanza. Among them the ghosts of dead kings were placed on an equality
with the gods and received the same honour and worship; they foretold
events which concerned the State, and they advised the living king, warning
him when war was likely to break out. The king consulted them periodically,
visiting first one and then another of the temples in which the mortal
remains of his predecessors were preserved with religious care. But the
temple (malolo) of a king contained only his lower jawbone and his navel-
string (mulongo); his body was buried elsewhere.440
For curiously enough
the Baganda believed that the part of the body to which the ghost of a dead
man adheres above all others is the lower jawbone; wherever that portion
of his person may be carried, the ghost, in the opinion of these people, will
follow it, even to the ends of the earth, and will be perfectly content to
remain with it so long as the jawbone is [pg 168] honoured.441
Hence the
jawbones of all the kings of Uganda from the earliest times to the present
day have been preserved with the utmost care, each of them being
deposited, along with the stump of the monarch's navel-string, in a temple
specially dedicated to the worship of the king's ghost; for it is believed that
the ghosts of the deceased monarchs would quarrel if they shared the same
temple, the question of precedence being one which it would be very
difficult for them to adjust to their mutual satisfaction.442
All the temples of
the dead kings stand in the district called Busiro, which means the place of
the graves, because the tombs as well as the temples of the departed
potentates are situated within its boundaries. The supervision of the temples
and of the estates attached to them was a duty incumbent on the Mugema
or earl of Busiro, one of the few hereditary chiefs in the country. His
principal office was that of Prime Minister (Katikiro) to the dead kings.443
When a king dies, his body is sent to Busiro and there embalmed. Then it is
laid to rest in a large round house, which has been built for its reception on
the top of a hill. This is the king's tomb. It is a conical structure supported
by a central post, with a thatched roof reaching down to the ground. Round
the hut a high strong fence of reeds is erected, and an outer fence encircles
48. f
f
f
the whole at some distance lower down the hill. Here the body is placed on
a bedstead; the sepulchral chamber is filled with bark cloths till it can hold
no more, the mainpost is cut down, and the door of the tomb closed, so
that no one can enter it again. When that was done, the wives of the late
king used to be brought, with their arms pinioned, and placed at intervals
round the outer wall of the tomb, where they were clubbed to death.
Hundreds of men were also killed in the space between the two fences, that
their ghosts might wait on the ghost of the dead king in the other world.
None of their bodies were buried; they were left to rot where they fell. Then
the gates in the fences were closed; and three chiefs [pg 169] with their
men guarded the dead bodies from the wild beasts and the vultures. But the
hut in which the king's body reposed was never repaired; it was allowed to
moulder and fall into decay.444
Five months later the jawbone of the royal corpse was removed in order to
be fashioned into an effigy or representative of the dead king. For this
purpose three chiefs entered the tomb, not through the door, but by cutting
a hole through the wall, and having severed the head from the body they
brought it out, carefully filling up the hole in the wall behind them, replacing
the thatch, and securing the gates in the fence. When the jawbone had
been removed by a chief of the Civet clan, the skull was sent back to Busiro
and buried with honour near the mouldering tomb. In contrast to the
neglect of the tomb where the royal body lay, the place where the skull was
buried was kept in good repair and guarded by some of the old princesses
and widows. As for the jawbone, it was put in an ant-hill and left there till
the ants had eaten away all the flesh. Then, after it had been washed in
beer and milk, it was decorated with cowry-shells and placed in a wooden
vessel; this vessel was next wrapt in bark cloths till it assumed a conical
shape, about two and a half feet high by a foot and a half broad at the
base. This conical packet, decorated on the outside with beads, was treated
as an image of the deceased king or rather as if it were the king himself in
life, for it was called simply “The King.” Beside it was placed the stump of
the king's navel-string, similarly wrapt in bark cloths and decorated, though
not made up into a conical shape.445
The reason for preserving both the
jawbone and the navel-string was that the ghost of the king was supposed
to attach itself to his jawbone, and the ghost of his double to his navel-
string. For in the belief of the Baganda every person has a double, namely,
the afterbirth or placenta, which is born immediately after him and is
regarded by the [pg 170] people as a second child. Now that double has a
49. f
ghost of its own, which adheres to the navel-string; and if the person is to
remain healthy, it is essential that the ghost of his double should be
carefully preserved. Hence every Baganda man and woman keeps his or her
navel-string wrapt up in bark cloth as a treasure of great price on which his
health and prosperity are dependent; the precious little bundle is called his
Twin (mulongo), because it contains the ghost of his double, the afterbirth.
If that is deemed necessary for everybody, much more is it deemed
essential for the welfare of the king; hence during his life the stump of his
navel-string is kept, as we saw,446
by one of the principal ministers of state
and is inspected by the king himself every month. And when his majesty has
departed this life, the unity of his spirit imperatively demands that his own
ghost and the ghost of his double should be kept together in the same
place; that is why the jawbone and the navel-string of every dead king are
carefully preserved in the same temple, because the two ghosts adhere
respectively to these two parts of his person, and it would be unreasonable
and indeed cruel to divide them.447
The two ghosts having been thus safely lodged in the two precious parcels,
the next thing was to install them in the temple, where they were to enter
on their career of beneficent activity. A site having been chosen, the whole
country supplied the labour necessary for building the temple; and ministers
were appointed to wait upon the dead king. The officers of state who had
held important posts during his life retained their titles and continued to
discharge their duties towards their old master in death. Accordingly houses
were built for them near the temple. The dowager queen also took up her
residence at the entrance to the temple enclosure, and became its principal
guardian. Many also of the king's widows of lower rank were drafted off to
live inside the enclosure and keep watch over it. When the queen or any of
these widows died, her place was supplied by another princess or a [pg
171] woman of the same clan; for the temple was maintained in perpetuity.
However, when the reigning king died, the temple of his predecessor lost
much of its importance, though it was still kept up in a less magnificent
style; indeed no temple of a dead king was allowed to disappear
altogether.448
Of all the attendants at the temple the most important
probably was the prophet or medium (mandwa), whose business it was
from time to time to be inspired by the ghost of the deceased monarch and
to give oracles in his name. To this holy office he dedicated himself by
drinking a draught of beer and a draught of milk out of the dead king's
skull.449
50. f
f
The temple consecrated to the worship of a king regularly stood on a hill.
The site was generally chosen by the king in his life, but sometimes his
choice was set aside by his successor, who gave orders to build the temple
in another place.450
The structure was a large conical or bee-hive-shaped
hut of the ordinary pattern, divided internally into two chambers, an outer
and an inner. Any person might enter the outer chamber, but the inner was
sacred and no profane person might set foot in it; for there the holy relics of
the dead king, his jawbone and his navel-string, were kept for safety in a
cell dug in the floor, and there, in close attendance on them, the king's
ghost was believed to dwell. In front of the partition which screened this
Holy of Holies from the gaze of the multitude there stood a throne, covered
with lion and leopard skins and fenced off from the rest of the sacred edifice
by a glittering rail of brass spears, shields, and knives. A forest of poles,
supporting the roof, formed a series of aisles in perfect line, and at the end
of the central nave appeared, like the altar of a Christian church, the throne
in all its glory. When the king's ghost held a reception, the holy relics, the
jawbone and the navel-string, each in its decorated wrappings, were
brought forth and set on the throne; and every person who entered the
temple bowed to the ground [pg 172] and greeted the jawbone in an
awestruck voice, for he regarded it as the king in person. Solemn music
played during the reception, the drums rolling and the women chanting,
while they clapped their hands to the rhythm of the songs. Sometimes the
dead king spoke to the congregation by the voice of his prophet. That was a
great event. When the oracle was about to be given to the expectant
throng, the prophet stepped up to the throne, and addressing the spirit
informed him of the business in hand. Then he smoked one or two pipes,
and the fumes bringing on the prophetic fit, he began to rave and to speak
in the very voice and with the characteristic turns of speech of the departed
monarch, for the king's spirit was now in him. This message from the world
beyond the grave was naturally received with rapt attention. Gradually the
fit of inspiration passed: the voice of the prophet resumed its natural tones:
the spirit had departed from him and returned to its abode in the inner
room. Such a solemn audience used to be announced beforehand by the
beating of the drums in the early morning, and the worshippers brought
with them to the temple offerings of food for the dead king, as if he were
still alive.451
But the greatest day of all was when the reigning king visited the temple of
his father. This he did as a rule only once during his reign. Nor did the
51. f
f
people approve of the visits being repeated, for each visit was the signal for
the death of many. Yet, attracted by a painful curiosity, crowds assembled,
followed the monarch to the temple, and thronged to see the great
ceremony of the meeting between the king and the ghost of his royal father.
The sacred relics were displayed: an old man explained them to the
monarch and placed them in his hands: the prophet, inspired by the dead
king's spirit, revealed to the living king his destiny. The interview over, the
king was carried back to his house. It was on the return journey that he
always gave, suddenly and without warning, the signal of death. Obedient
to his [pg 173] orders the guards rushed upon the crowd, captured
hundreds of spectators, pinioned them, marched them back to the temple,
and slaughtered them within the precincts, that their ghosts might wait on
the ghost of the dead king.452
But though the king rarely visited his father's
ghost at the temple, he had a private chapel for the ghost within the vast
enclosure of the royal residence; and here he often paid his devotions to the
august spirit, of whom he stood greatly in awe. He took his wives with him
to sing the departed monarch's praise, and he constantly made offerings at
the shrine. Thither, too, would come the prophet to suck words of wisdom
from the venerable ghost and to impart them to the king, who thus walked
in the counsel of his glorified father.453
In Kiziba, a district of Central Africa on the western side of Lake Victoria
Nyanza, the souls of dead kings become ruling spirits; temples are built in
their honour and priests appointed to serve them. The people are composed
of two different races, the Bairu, who are aboriginals, and the Bahima, who
are immigrants from the north. The royal family belongs to the Bahima
stock. In his lifetime the king's person is sacred; and all his actions,
property, and so forth are described by special terms appropriated to that
purpose. The people are divided into totemic clans: the totems (muziro) are
mostly animals or parts of animals: no man may kill or eat his totem animal,
nor marry a woman who has the same totem as himself. The royal family
seems to have serpents for their totem; after death the king's soul lives in a
serpent, while his body is buried in the hut where he died. The people
revere a supreme god named Rugaba, who is believed to have created man
and cattle; but they know little about him, and though they [pg 174]
occasionally pray to him, particularly in the case of a difficult birth, he has
no priests and receives no sacrifices. The business of the priests is to act as
intermediaries, not between God and man, but between men and the
spirits. The spirits are believed to have been formerly kings of the world.
52. Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.
More than just a book-buying platform, we strive to be a bridge
connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.
Join us on a journey of knowledge exploration, passion nurturing, and
personal growth every day!
ebookbell.com